JP4719854B2 - Moving picture encoding apparatus, moving picture encoding program, moving picture decoding apparatus, and moving picture decoding program - Google Patents

Description

  The present invention relates to a moving image encoding device, a moving image encoding program, a moving image decoding device, and a moving image decoding program, and in particular, when encoding a moving image, a condition that satisfies the Poisson equation is defined as a boundary condition. The present invention also relates to a moving image encoding device, a moving image encoding program, a moving image decoding device, and a moving image decoding program that are derived based on a source model suitable for moving images and apply the results to encoding and decoding.

  In recent years, various moving image distribution services via digital broadcasting and networks have become common, and in order to record more high-quality and high-quality moving images, further improvement in the coding efficiency of moving images is desired. ing.

  Generally, video information to be transmitted is often multimedia information created by combining digitized voice, image, and other data. The information amount of such video information is generally 1 to 2 bytes when the data is simply digitized and converted. In the case of voice, the telephone quality is 64 kbits per second. In the case of a moving image, even if it is SD analog terrestrial television broadcast quality, an information amount of 100 Mbits per second or more is required. It is not practical to handle such a large amount of digital information as it is in view of the cost of using the current transmission path and recording medium.

  Therefore, an encoding technique for compressing such an enormous amount of information is currently being actively researched and developed. MPEG (Moving Picture Experts Group) is a representative example of video encoding technology. MPEG is an international standard for moving image data encoding technology. At present, the MPEG standard mainly supports MPEG-1 for storage media of about 1.5 Mbps such as a video CD, and various applications such as communication and broadcasting such as digital broadcasting. -2, MPEG-4 that can realize more advanced multimedia contents by object coding and control thereof, and MPEG-7 for content description are defined. By applying such an encoding technique, it is possible to efficiently encode video data having an enormous amount of information and compress the information amount to about 1/20 to 1/40.

  In addition, some of the new coding methods such as MPEG-4 AVC (hereinafter referred to as AVC or H.264) that have greatly improved the coding efficiency of the conventional coding technology have been introduced. Thus, more effective information encoding can be performed (see, for example, Non-Patent Document 1 and Non-Patent Document 2).

  MPEG is configured to perform more effective compression by combining a plurality of technologies. 29 and 30 are block diagrams showing an example of a conventional moving picture encoding apparatus and decoding apparatus based on MPEG. FIG. 29 is a block diagram showing a basic configuration of a conventional moving picture encoding apparatus based on the AVC. An input code 101 as an input image is supplied to a subtractor 102, where the motion compensation (selected by the switch 113) ( MC) The difference between the prediction picture generated by decoding and prediction by the predictor 112 or the intra predictor 110 is taken to reduce redundant components included in the time direction, and then supplied to the orthogonal transformer 103. The

  Here, AVC has three prediction modes: intra prediction (Intra) coding, forward prediction (Predictive) coding, and bi-predictive coding. In the intra prediction encoding, after using the in-plane information and the in-plane prediction information by the intra predictor 110 without using the output of the motion compensation predictor 112, the transform by the orthogonal transformer 103 described later is performed. . What is encoded in this mode is called an I picture. An I picture can be decoded without depending on other pictures at the time of decoding. In AVC, normally, one picture can include a unit of a plurality of slices. However, for the sake of simplicity, it is assumed that one picture is composed of only one slice.

  In forward predictive coding, a motion-compensated predictor 112 compensates for a previously coded picture and performs prediction for the current coding target picture. A transform by the orthogonal transformer 103 is performed on the difference picture between the prediction picture generated by this prediction and the input code 101. For the orthogonal transform performed by the orthogonal transformer 103, DCT (Discrete Cosine Transform) is used in MPEG-1, -2, and 4. Orthogonal transformation can also be performed using other transformation bases such as Hadamard bases and wavelet bases. In AVC, orthogonal transform is performed using integer precision DCT and Hadamard transform. What is encoded in the forward prediction mode is called a P picture. P pictures depend on other pictures and cannot be decoded independently like I pictures, but can have a higher compression rate than I pictures.

  In bi-predictive coding, the motion compensated predictor 112 compensates not only for past and future bi-directional, but also for any two encoded frames, and performs prediction for the current coding target picture. A transform by the orthogonal transformer 103 is performed on the difference picture between the reference picture generated by this prediction and the input code 101. A picture encoded in the bi-prediction mode is called a B picture. The B picture depends on other pictures in the past and the future, so it cannot be decoded independently like the I picture and P picture, but the compression rate can be further increased than the I picture and P picture.

  Each frame and each slice are subjected to prediction processing for each MB (macroblock) of 16 pixels in the horizontal direction and 16 pixels in the vertical direction as a predetermined rectangular area. In addition, prediction processing may be performed by dividing into smaller block units or sub-block units. Here, the discussion will be continued on the assumption that processing is performed in units of blocks. The prediction direction differs depending on the I picture, P picture, and B picture. It is the I picture that encodes all MBs independently. The P picture has two modes: a mode in which encoding is performed by prediction from a past picture, and a mode in which the MB is independently encoded without prediction. In addition, there are four modes in which the MB is independently encoded without prediction, prediction from the future, prediction from the past, prediction from both, and prediction.

  In order to identify the region where the error is minimum for each MB in the moving image frame to be encoded by the motion estimation (ME) unit 111 from the reference image frame, A predetermined pattern matching is performed with an accuracy of 1/4 pixel to generate motion information. Here, the motion information is composed of at least motion vector information for identifying a spatial position up to an area corresponding to each MB detected by pattern matching and reference picture information for identifying a reference picture. . Motion compensation is performed by the motion compensator 112 generating a predicted image frame from the reference image frame based on the motion information estimated by the motion estimator 111.

  As described above, the orthogonal transform is performed on the differential picture output from the subtractor 102 by the orthogonal transformer 103. This orthogonal transform uses DCT in MPEG-1, -2 and -4, but uses integer precision DCT and Hadamard transform in AVC. The DCT is an orthogonal transform that discretely transforms an integral transform based on a cosine function into a finite space. In MPEG-1, -2, and -4, two-dimensional DCT is performed on a DCT block of 8 pixels in the horizontal direction and 8 pixels in the vertical direction, which is a predetermined rectangular area by dividing MB into four. In AVC, two-dimensional integer precision DCT and Hadamard transform are performed on a block of 4 pixels in the horizontal direction and 4 pixels in the vertical direction. In general, since a video signal has many low-frequency components and few high-frequency components, the coefficients can be concentrated on the low-frequency components by performing DCT, and the subsequent quantizer 104 can efficiently reduce the amount of information. .

  The DCT coefficient obtained by the orthogonal transformer 103 is quantized by the quantizer 104. Quantization is a quantization matrix of MPEG-1, -2 and -4, a value obtained by weighting the two-dimensional frequency of 8 pixels in the horizontal direction and 8 pixels in the vertical direction with visual characteristics, and 4 pixels in the horizontal direction and 4 pixels in the vertical direction in AVC. A value obtained by multiplying a value obtained by weighting the two-dimensional frequency with a visual characteristic and a value called a quantization scale for multiplying the whole by a scalar is used as a quantization value, and the DCT coefficient is divided by the quantization value. When inverse quantization is performed by the decoder, a value close to the original DCT coefficient is obtained by multiplying by the quantized value.

  The data quantized by the quantizer 104 is encoded by the entropy encoder 116. In general, entropy coding is a VLC (Variable Length Code) device, and performs variable length coding on quantized data. In variable-length coding, in general, zigzag scanning is performed from the low range to the high range of the DCT coefficient, the run length of zero and the effective coefficient value are regarded as one event, and a code with a short code length is assigned from the one with the highest appearance probability. Huffman encoding is performed. Further, the entropy encoder 116 can compress higher information by using arithmetic coding instead of Huffman coding. In AVC, CAVLC (Context-Adaptive Variable Length Coding) and CABAC (Context-Adaptive Binary Arithmetic Coding) are used for this entropy coding. Increases efficiency.

  The entropy-encoded data is output as output code (encoded data) 117 at a predetermined transfer rate. FIG. 29 can also include a code amount controller for accumulating encoded data and controlling the code amount. The code amount for each macroblock of the encoded data to be output is notified to a code amount controller not described in FIG. 29, and the error code amount between the notified code amount and the target code amount is quantized. 104 is notified. The quantizer 104 can control the code amount by adjusting the quantization scale based on the notified error code amount information.

  The quantized data output from the quantizer 104 is inversely quantized by the inverse quantizer 105. Further, the inverse orthogonal transformer 106 performs inverse orthogonal transform using a transform base corresponding to the orthogonal transformer 103. The signal output from the inverse orthogonal transformer 106 is added to the frame from the motion compensator 112 or the intra predictor 110 selected by the switch 113 in the adder 107, and then supplied to the intra predictor 110. On the other hand, the deblocking filter 108 performs deblocking processing for reducing block distortion generated at the time of encoding, and then stores it in the frame memory 109. The frame stored in the frame memory 109 is used as a reference frame (reference picture) for obtaining a predicted picture in the motion estimator 111 and the motion compensator 112.

  The output code 117 output in this way is distributed as an encoded bit stream by transmission through a storage medium or a network, and is reproduced by a terminal device. In the case of reproduction by the terminal device, the distributed coded bit stream is reproduced after being decoded by the decoding device of the terminal device. FIG. 30 shows a configuration example of an AVC decoding apparatus as an example of a conventional decoding apparatus, which will be briefly described below.

  In FIG. 30, when an encoded bit stream is input to a decoding apparatus as an input code 201, it is separated into pieces of information such as entropy-encoded motion vector information and texture information, and each piece of information is entropy decoder 202. Will be notified. The entropy decoder 202 decodes the entropy-encoded texture information and notifies the inverse quantizer 203. In addition, the entropy decoder 202 decodes the entropy-encoded motion vector information and notifies the motion compensation predictor 206. In general, entropy decoding is an inverse VLC unit that performs variable length decoding on demultiplexed data. In AVC, variable length decoding is performed by inverse CAVLC or inverse CABAC.

  The inverse quantizer 203 calculates a value close to the original orthogonal transform coefficient by multiplying the entropy-decoded texture information by the quantized value, and then supplies it to the inverse orthogonal transformer 204. The inverse orthogonal transformer 204 obtains original frame information or inter-frame difference information by performing inverse orthogonal transform on the supplied orthogonal transform coefficient. This inverse orthogonal transform uses IDCT (Inverse Discrete Cosine Transform) in MPEG-1, -2 and -4. AVC uses integer precision IDCT and Hadamard transform.

  The frame information or interframe difference information obtained by the inverse orthogonal transformer 204 is supplied to the adder 205 and added with the predicted frame selected by the switch 207 from the motion compensation predictor 206 or the intra predictor 208. Later, it is supplied to the intra predictor 208 and is subjected to deblocking processing for reducing block distortion by the deblocking filter 209 and then stored in the frame memory 210.

  The motion compensator 206 uses the motion vector information obtained from the entropy decoder 202 and the reference frame obtained from the frame memory 210 to shift the corresponding region based on the motion vector information and perform prediction. Create a prediction frame. This predicted frame is supplied to the adder 205 via the switch 207. In this way, the input code 201 is decoded by the decoding device of the terminal device, and the video information is reproduced.

  In addition, in recent years, JPEG (Joint Photographic Experts Group) encoding, which is known as a natural image encoding scheme, is further improved in encoding efficiency by using multiple harmonic local orthogonal transform, particularly multiple harmonic local cosine transform. There has been proposed a method for trying (see, for example, Non-Patent Document 3). In this multi-harmonic local cosine transform method, the concept of the Poisson equation is introduced to obtain an estimated signal for the original signal in the block, and a residual signal that is a difference between the original signal and the obtained estimated signal is generated. It is possible to quickly converge DCT coefficients having a higher frequency when introducing, and as a result, it is possible to improve the encoding efficiency.

  FIG. 31 is a conceptual diagram for illustrating a basic concept of this multiple harmonic local cosine transform method. FIG. 31 shows a particular block in an original image frame in which attention is paid to a particular line in the block, and the state of the original signal on the noticed line and the boundary of the original signal on the block boundary is shown in FIG. Shows how it changes.

  With respect to the original signal 301 shown in FIG. 31A, the inclinations 302 and 303 of the image signal at both ends of the original signal corresponding to the block boundaries 304 and 305 are obtained. The slopes 302 and 303 of the image signals at both ends are used as boundary conditions. In the multiple harmonic local cosine transform method, a predetermined source model that defines the behavior of a signal in a block is applied, and an estimated signal is generated from the signal in the block based on the boundary condition of the block.

  Here, the estimation signal 306 that satisfies the Poisson equation is generated using the gradients of the image signals at both ends of the original signal as shown in FIG. As the predetermined source model, it is desirable to adopt a model that can estimate a signal closest to the original signal 301 under boundary conditions. In general, in the case of a one-dimensional signal as shown in FIG. 31, it can be easily obtained by applying a quadratic function as a predetermined source model, but here it is not particularly limited to this model. Note that there is no. In this way, by introducing a predetermined source model, an estimated signal in the block is generated analytically without spending an enormous amount of computation and solving the Poisson equation numerically.

  After generating the estimated signal 306 based on the boundary condition as shown in FIG. 31 (b), the difference between the original signal 301 and the estimated signal 306 is obtained to obtain the residual as shown in FIG. 31 (c). A signal 307 is generated. The residual signal 307 thus obtained is subjected to normal orthogonal transformation, such as DCT, quantization, and entropy coding, thereby completing the natural image coding using the multiple harmonic local cosine transformation method. .

  FIG. 32 and FIG. 33 are diagrams for showing what difference is produced from the normal DCT by using the multiple harmonic local cosine transform method. FIG. 32 shows an example in which normal DCT is performed on the original signal shown in FIG. In FIG. 32, the horizontal axis indicates x in the x direction of the image signal when attention is paid only to the x direction, and the vertical axis indicates the value (luminance value) of each pixel of one line of the image signal. In FIG. 32, the waveform is shown as a continuous function for convenience, but it is actually a discrete value (the same applies to FIG. 33).

  Since the DCT type normally uses DCT-II, when DCT is performed on a finite original signal divided by blocks as shown in FIG. On one side, here, the original signal is evenly connected with the right side as an axis, and a periodic waveform signal of the original signal as shown in FIG.

  By using such a periodic waveform signal, it can be expressed by a DCT series as shown in FIG. Usually, as indicated by dotted line portions 401 to 403 in FIG. 32B, when the original signal is evenly connected, a portion where the smoothness of the signal is not always sufficient is generated. Because of this influence, the DCT coefficients, which are the amplitudes of the DCT series shown in FIG. 32C, are mixed with signal components that are different from the signal components inherent to the original signal due to the even connection of the signals. Even in a high frequency component region as indicated by the dotted line portion 404, the DCT coefficient is not necessarily sufficiently converged. Accordingly, when predetermined quantization and entropy coding is performed on such a DCT coefficient, a corresponding code amount is generated.

  Next, FIG. 33 shows a case where the multiple harmonic local cosine transform method is used. After generating the estimated signal 306 as shown in FIG. 31 (b) using this method, the difference from FIG. 31 (a) is taken to obtain the remaining signal as shown in FIGS. 31 (c) and 33 (a). A difference signal 307 is obtained. By using this residual signal 307, as shown in FIG. 33 (a), the residual signal 307 is evenly connected with the right side of the signal as an axis as shown in FIG. 32 (a). A periodic waveform signal of the residual signal as shown in FIG.

  The obtained periodic waveform signal of the residual signal is a waveform of a portion corresponding to the periodic waveform signal of the original signal shown in FIG. 32B in the dotted line portions 501 to 503 of FIG. As a result, it is possible to improve smoothness and to prevent unnecessary signal components that are not originally included in the original signal from being mixed. Therefore, when the DCT is performed by regarding the periodic waveform signal of the residual signal shown in FIG. 33B as a discrete periodic waveform and the spectrum thereof is expressed, a DCT coefficient sequence as shown by I in FIG. Residual orthogonal transform coefficients (here, residual DCT coefficient information) are obtained. The residual DCT coefficient information indicated by I in FIG. 33C is compared with the DCT coefficient sequence indicated by II in FIG. 33C when normal DCT is performed, and the dotted line portion 505 in FIG. The DCT coefficient is sufficiently suppressed in the high frequency component region as shown, and the state is biased toward the DCT coefficient having a lower frequency component.

  Therefore, by performing predetermined quantization and entropy coding on the DCT coefficients using the multiple harmonic local cosine transform method in this way, the amount of codes generated can be suppressed as compared with the case where coding by normal DCT is performed. This leads to an improvement in coding efficiency.

  The above is the basic concept when the multiple harmonic local cosine transform method is used. Here, in order to briefly explain the basic concept of the multiple harmonic local cosine transform method, an estimated signal is generated based on boundary conditions of the original signal, and a DCT is performed after generating a residual signal based on the difference from the original signal. It was shown that it will be realized.

  In addition, when DCT is performed on the residual signal by performing DCT on the original signal in the block and the estimated signal from the boundary condition to obtain respective DCT coefficients, A method for obtaining similar results has been conventionally disclosed (for example, see Non-Patent Document 3).

  Below, although it is simple, the method shown by the nonpatent literature 3 is demonstrated. First, an image frame signal is captured as each block shown in FIG.

図３４のΩ^{（０，０）}が現在処理対象となっているブロックであるものとする。また、各ブロック内の各画素は、数２で表現される。

It is assumed that Ω ^{(0, 0) in} FIG. 34 is a block currently being processed. In addition, each pixel in each block is expressed by Equation 2.

まず、各ブロックの画像信号に対して直交変換を行う。ここでは数３に示すようなＤＣＴを行う。

First, orthogonal transformation is performed on the image signal of each block. Here, DCT as shown in Equation 3 is performed.

次に、境界条件である画像信号の傾きを周波数領域で議論することができるように、ＤＣＴ級数で表現する。ここでは数４に示した境界条件のＤＣＴ級数展開の式を利用して画像信号の傾きをｇとした場合に数４に示した式中のＧ_ｋのようなＤＣＴ係数で表現する。

Next, the gradient of the image signal, which is a boundary condition, is expressed by a DCT series so that it can be discussed in the frequency domain. Here, when the gradient of the image signal is set to g using the DCT series expansion formula of the boundary condition shown in Formula 4, it is expressed by a DCT coefficient such as G _{k in} the formula shown in Formula 4.

また、ポアソン方程式の概念を導入してブロック内の原信号に対する推定信号を求める。ここで、ポアソン方程式は、処理対象ブロックΩｊにおいて、ブロック内の推定信号ｕのラプラシアンであるΔｕ_ｊがソース項Ｋ_ｊとの間において、次式が成立する方程式である。

In addition, the concept of the Poisson equation is introduced to obtain an estimated signal for the original signal in the block. Here, the Poisson equation is an equation in the processing target block Ωj, in which Δu _j that is the Laplacian of the estimated signal u in the block is between the source term K _j and the following equation is established.

Δu _j = K _j
The estimated signal u in the block can be expressed as shown in Equation 6 using Neumann boundary conditions and DCT series notation. This can be obtained by adding the DCT series expansion components of the estimation signal from each boundary as shown in Equation 5.

このような推定信号ｕに対してＤＣＴを行うことで、次式で表される推定信号Ｕを得ることができる。

By performing DCT on such an estimated signal u, an estimated signal U represented by the following equation can be obtained.

_{Uk1, k2} = _Gk1 ⁽¹⁾ _{ηk1, k2} + _Gk1 ⁽²⁾ η ^* _{k1, k2} + _Gk2 ⁽³⁾ _{ηk2, k1} + _Gk2 ⁽⁴⁾ η ^* _{k2, k1} (1)
However, in the equation (1), η _{k, m} and η ^* _{k, m} are expressed by the following equations.

ここで、数７に示した式から、η及びη^*はブロック内の位置情報のみに依存し、ブロック内の画像信号には依存しないことから、予め一意に計算して求めておくことができる。従って、推定信号Ｕを求めるためには、傾きのＤＣＴ係数情報Ｇ_ｋを求めることが重要となる。

Here, from the equation shown in Equation 7, since η and η ^* depend only on the position information in the block and do not depend on the image signal in the block, they can be uniquely calculated in advance. . Therefore, in order to obtain the estimated signal U, it is important to obtain the DCT coefficient information G _k slope.

Here, in order to enable the determination by using the DCT coefficient information after the DCT and G _k with respect to the original signal in a block, when determining the tilt of the image signal in the block boundary is the boundary condition The slope between the blocks is calculated using the average value of the signal components in each direction in the block as represented by the equation (8) as a representative value, and g is obtained by approximating this slope as the slope at the block boundary.

続いて、この近似されたｇに対してＤＣＴを行うことでＧを求める。ここで、Ｇを求める際に数８に示した式を考慮して数９に示した式変形を行うことで、ＧをＦ、つまりブロック内のＤＣＴ係数情報を利用して求めることが可能となる。

Subsequently, G is obtained by performing DCT on the approximated g. Here, when obtaining G, the equation shown in Equation 9 is considered in consideration of the equation shown in Equation 8, so that G can be obtained using F, that is, DCT coefficient information in the block. Become.

Supervised by Jun Okubo, Shinya Tsuno, Yoshihiro Kikuchi, Teruhiko Suzuki “H.264 / AVC textbook”, Impress, Inc., August 2004 Co-authored by Sadayasu Ono, Atshimichi Murakami, and Kotaro Asai, “High-efficiency coding of moving images”, Ohm Corporation, April 2005 Masaru Yamatani, Naoki Saito, “Improvement of DCT-based Compression Algorithms Using Poisson's Equation”

  By the way, in the conventional video coding, DCT is often used for orthogonal transform. MPEG-4 AVC uses integer DCT. The orthogonal transform used in the conventional moving image encoding is suitable for encoding natural images. In moving image coding, generally, a difference image generated between a prediction image in which signal continuity is not always maintained in a spatial direction by motion estimation / compensation in units of blocks and a picture to be encoded is mainly used. Are subject to orthogonal transformation. Since this difference image is considered to have a property different from that of a natural image, it is not necessarily effective to use a normal DCT for orthogonal transform.

When transforming a block area using conventional orthogonal transform, period expansion is performed in the frequency domain with respect to the block area to be encoded, so unnecessary high frequency components at the block boundary when viewed as an entire image Will occur. In addition, since such unnecessary high frequency components are also to be encoded, an unnecessarily necessary code amount is generated, resulting in a decrease in encoding efficiency.

  Further, in DCT, when transforming a block area which is an analysis unit, orthogonal transformation is performed by regarding this block area as a periodic waveform. In other words, as described with reference to FIG. 32, since the period extension is performed on the block region, discontinuous points are generated at the block boundary without smooth connection. That is, when the entire image is viewed at this block boundary, an unnecessary high frequency component is generated, which becomes an obstacle in encoding.

  In addition, by applying a multi-harmonic local cosine transform method that attempts to further improve the encoding efficiency as compared with JPEG as a conventional method, the problem of encoding natural images is improved. However, this multi-harmonic local cosine transform method is assumed to be applied to a general natural image, and in order to suppress the occurrence of block noise, encoding depending on the low frequency component between blocks is performed.

  In order to suppress the occurrence of such block noise, encoding deterioration at a low bit rate is achieved by encoding while overlapping the encoding target, such as MDCT (modified DCT) and LOT (overlapping orthogonal transform). There are ways to reduce this. However, generally used video coding cannot always maintain signal continuity in the spatial direction due to motion estimation / compensation in units of blocks, and sufficient coding efficiency is achieved when coding is performed while overlapping. May not be obtained.

  The conventional multiple harmonic local cosine transform method assumes that continuity between blocks of an image frame to be encoded is maintained as in a general natural image. Here, the application of the conventional multi-harmonic local cosine transform method is considered for general moving picture coding with motion estimation and motion compensation prediction. When the prediction accuracy of motion estimation / motion compensation is not sufficiently high, a predicted image frame in which continuity is not always maintained between adjacent blocks in the frame is generated.

  A difference image frame including a residual component is generated from the difference between the predicted image frame and the original image frame to be encoded. Even in this difference image frame, continuity is always maintained between adjacent blocks. However, there is a case where the block boundary becomes discontinuous. When the conventional multiharmonic local cosine transform method is applied to such a difference image frame, the correlation between the blocks is not necessary even though it is necessary to perform encoding so as to keep the discontinuity between the blocks. Encoding processing is performed on the assumption that the continuity is high and sufficient encoding efficiency may not be obtained as compared with a natural image.

  Also, if the amount of code required to reproduce the discontinuity between blocks at the time of encoding has been reduced by processing such as quantization, the decoding side performs decoding assuming continuity between blocks Therefore, discontinuity between blocks cannot be reproduced, and a problem arises in that this influence is propagated as an error by motion compensated prediction of moving picture decoding.

  In addition, when intra coding is performed on an input moving image, coding is generally performed using a correlation in a spatial direction of a picture to be coded. However, the spatial correlation in the picture to be encoded is not necessarily composed of continuous signal components throughout the picture. For example, it is assumed that a picture to be encoded has a predetermined object and background. In such a picture, the signal components in the object are highly correlated with each other, and the signal components are also highly correlated with each other in the background portion, but the correlation between the signal components in the object and the signal components in the background portion is low. Considering the case, a sudden signal change, that is, a signal component discontinuity occurs at the boundary between the object and the background. In order to express such discontinuous and suddenly changing signal components, it is generally necessary to have a wide range from low frequency signal components to high frequency signal components, and the bias of the signal components is reduced. When normal intra coding is performed on a picture including components, there is a problem that sufficient coding efficiency cannot be obtained.

  Furthermore, in the intra prediction used in the conventional video coding and decoding by AVC, the block closest to the encoding target block is selected from the prediction blocks generated by the predetermined direction and the predetermined weighting. Improvement of coding efficiency in intra coding is attempted. However, since this prediction block is generated with a limited direction and a limited weight, when the encoding target block includes a contour portion characterizing the shape of the image or includes a complicated change. Is not necessarily a good approximation of the block to be encoded, and sufficient encoding efficiency may not be obtained.

  The present invention has been made in view of the above points, and is an encoding method using a multiharmonic local cosine transform method by canceling the discontinuity of the boundary portion in the image signal included in the block to be encoded. Another object of the present invention is to provide a moving image encoding device, a moving image encoding program, a moving image decoding device, and a moving image decoding program that can perform decoding.

  Another object of the present invention is to provide a moving picture coding apparatus, a moving picture coding program, a moving picture decoding apparatus, and a moving picture decoding program that can improve coding efficiency.

  Furthermore, another object of the present invention is to provide a moving image that can improve the phenomenon that discontinuity between blocks is propagated by motion compensated prediction of moving image decoding when a multiple harmonic local cosine transform method is applied. To provide an encoding device, a moving image encoding program, a moving image decoding device, and a moving image decoding program.

  Furthermore, another object of the present invention is to obtain a sufficient encoding efficiency when a normal intra coding is performed on a picture including a signal component with discontinuous and rapid change, and at a higher speed. An object of the present invention is to provide a moving image encoding device, a moving image encoding program, a moving image decoding device, and a moving image decoding program that can perform encoding and decoding.

  Still another object of the present invention is to provide a moving picture coding apparatus capable of obtaining sufficient coding efficiency even if the encoding target block includes a contour portion characterizing the shape of the image or includes a complicated change. Another object is to provide a moving picture encoding program, a moving picture decoding apparatus, and a moving picture decoding program.

In order to achieve the above object, the moving picture coding apparatus according to the first aspect of the present invention subdivides one screen area of the inputted moving picture signal to be coded in units of a rectangular area having a predetermined number of pixels. A code that generates a difference image signal that is a difference signal between a prediction image signal generated from a locally decoded image signal and a moving image signal to be encoded, with a rectangular area as a processing unit, and encodes the difference image signal In the device
Orthogonal transformation at the boundary between the orthogonal transformation means for performing orthogonal transformation on the difference image signal using the rectangular area as a processing unit and outputting orthogonal transformation coefficient information, and another rectangular area adjacent to the rectangular area to be encoded When generating approximate orthogonal transform coefficient information obtained by orthogonally transforming the estimation signal in the rectangular region to be encoded so as to satisfy the Poisson equation with the gradient of the moving image signal obtained from the transform coefficient information as a boundary condition , In order to approximate the generated approximate orthogonal transform coefficient information to the approximate orthogonal transform coefficient information obtained by the inverse multiple harmonic local orthogonal transform performed in the local decoding process, the orthogonal transform coefficient information obtained from the orthogonal transform means is quantized. based on the quantization parameter estimates the orthogonal transform coefficient information subjected to inverse quantization after quantization, and generates an approximate orthogonal transform coefficient information with the estimated result to be used in, A multi-harmonic local orthogonal transform means for generating residual orthogonal transform coefficient information by taking a difference between the alternating transform coefficient information and the generated approximate orthogonal transform coefficient information, and the residual orthogonal transform coefficient information based on a predetermined quantization parameter And an entropy code that generates predetermined post-entropy encoded information by performing entropy encoding on the quantized residual orthogonal transform coefficient information output from the quantizing means An encoding unit, a multiplexing unit that generates an encoded bitstream by performing predetermined multiplexing on the post-entropy encoded information output from the entropy encoding unit according to a predetermined syntax structure, and a quantization unit Inverse quantization means for inversely quantizing the output residual orthogonal transform coefficient information after quantization based on a predetermined quantization parameter, and a rectangular to be decoded A rectangle to be decoded that satisfies the Poisson equation with the gradient of the moving image signal obtained from the residual orthogonal transform coefficient information after dequantization at the boundary with the other adjacent rectangular region as the boundary condition Approximate orthogonal transform coefficient information obtained by orthogonal transform for the estimation signal in the region is generated using the residual orthogonal transform coefficient information after dequantization generated by the dequantization means, and after the inverse quantization Inverse multi-harmonic local orthogonal transform means for decoding orthogonal transform coefficient information by combining residual orthogonal transform coefficient information and generated approximate orthogonal transform coefficient information, and orthogonal generated by inverse multi-harmonic local orthogonal transform means Inverse orthogonal transform means for generating a locally decoded image signal by performing inverse orthogonal transform on the transform coefficient information is characterized.

  In the present invention, a multiple harmonic local orthogonal transformation method is used, and the gradient of the moving image signal obtained from the orthogonal transformation coefficient information at the boundary between the rectangular area to be encoded and another adjacent rectangular area is used as the boundary condition. Approximate orthogonal transformation coefficient information obtained by orthogonal transformation is generated using the orthogonal transformation coefficient information obtained from the orthogonal transformation means for the estimated signal in the rectangular area to be encoded that satisfies the Poisson equation. Then, the residual orthogonal transform coefficient information is generated by taking the difference between the orthogonal transform coefficient information and the generated approximate orthogonal transform coefficient information, and the residual orthogonal transform coefficient information is quantized based on a predetermined quantization parameter. Thereafter, entropy coding is performed on the post-quantization residual orthogonal transform coefficient information to generate predetermined post-entropy coded information, and the multiplexing means further performs entropy coding. To generate an encoded bit stream by performing predetermined multiplexing according to a predetermined syntax structure with respect of later information.

  Here, the residual orthogonal transform coefficient information, which is the difference between the approximate orthogonal transform coefficient information orthogonally transformed with respect to the estimated signal and the orthogonal transform coefficient information, has a zero slope at the boundary of the rectangular area. The periodic waveform signal of the difference orthogonal transform coefficient information improves the smoothness of the waveform at the periodic connection part, and can prevent mixing of unnecessary signal components not included in the original signal. Accordingly, the residual orthogonal transform coefficient information can be biased to an orthogonal transform coefficient having a lower frequency component while suppressing a high frequency component.

  In order to achieve the above object, the moving picture coding apparatus according to the second invention is further subject to coding based on the orthogonal transform coefficient information generated by the orthogonal transform means in addition to the first invention. Analyzing the discontinuity of the moving image signal between the adjacent rectangular area and the adjacent surrounding rectangular area, and determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition , Boundary condition analysis means for generating first boundary condition correction information for specifying whether or not boundary condition correction is necessary, and inverse quantization if the first boundary condition correction information is not valid Analyzing the discontinuity of the video signal between the rectangular region to be decoded and the adjacent rectangular region based on the residual orthogonal transform coefficient information after inverse quantization by the means, and at least one Boundary conditions based on predetermined judgment conditions The second boundary condition correction information for determining whether correction is necessary and specifying whether the boundary condition needs to be corrected is generated, and the generated second boundary condition correction information or Boundary condition determination means that uses valid first boundary condition correction information as boundary condition determination information is added, and the multiple harmonic local orthogonal transform means includes the first boundary condition obtained by the boundary condition analysis means. Based on the correction information, the inclination of the moving image signal obtained from the orthogonal transformation coefficient information which is the boundary condition of each rectangular area is controlled, and the encoding target is such that the Poisson equation is satisfied based on the boundary condition of each rectangular area. Approximate orthogonal transformation coefficient information obtained by orthogonal transformation of the estimated signal in the rectangular area is generated using the orthogonal transformation coefficient information after the orthogonal transformation, and the orthogonal transformation coefficient information obtained by the orthogonal transformation is generated. The residual orthogonal transform coefficient information is generated by taking the difference from the approximate orthogonal transform coefficient information, and the demultiplex harmonic local orthogonal transform means is configured to detect the rectangular region from the boundary condition determination information acquired from the boundary condition determination means. When the discontinuity of the boundary part is specified and the target boundary part is discontinuous, the inclination of the moving image signal obtained from the residual orthogonal transform coefficient information is corrected and the rectangular area to be decoded A rectangular region that is to be decoded so as to satisfy the Poisson equation with the gradient of the moving image signal obtained from the residual orthogonal transform coefficient information after inverse quantization at the boundary with the other adjacent rectangular region as the boundary condition The approximate orthogonal transform coefficient information obtained by orthogonal transform is estimated using the residual orthogonal transform coefficient information after the inverse quantization generated by the inverse quantization means, and the residual after the inverse quantization is obtained. Orthogonal difference It is a means for decoding orthogonal transformation coefficient information by combining the transformation coefficient information and the generated approximate orthogonal transformation coefficient information.

  In this invention, the discontinuity of the moving image signal between the rectangular area to be encoded and the adjacent surrounding rectangular area is analyzed, and the boundary condition is corrected based on at least one predetermined determination condition. Since it is determined whether or not it is necessary and whether or not the boundary condition needs to be corrected, it is possible to correctly reproduce the signal discontinuity between adjacent rectangular regions. Here, the orthogonal transform coefficient information may be DCT coefficient information, and the approximate orthogonal transform coefficient information may be approximate DCT coefficient information, and residual DCT coefficient information may be generated by taking a difference between them.

  In order to achieve the above object, the third invention subdivides one screen area of the input moving image signal to be encoded in units of a rectangular area having a predetermined number of pixels, and processes the rectangular area. In an encoding device that performs intra coding as a unit, orthogonal transformation is performed using a rectangular area as a processing unit, and orthogonal transformation coefficient information for each rectangular area is output in advance over the entire screen area of the moving image signal to be coded. In order to analyze the discontinuity of the moving image signal existing in one screen area of the moving image signal to be encoded It is necessary to analyze the discontinuity of the moving image signal between the existing rectangular area and determine whether or not the boundary condition needs to be corrected based on a predetermined determination condition and to correct the boundary condition. Whether or not The boundary condition analysis means for generating the first boundary condition correction information for calculating the inclination of the moving image signal obtained from the orthogonal transformation coefficient information at the boundary between the rectangular area adjacent to the encoding target rectangular area Use the orthogonal transform coefficient information obtained from the orthogonal transform means and the approximate orthogonal transform coefficient information obtained by orthogonal transform for the estimated signal in the rectangular area to be encoded that satisfies the Poisson equation as the boundary condition. And then generating the residual orthogonal transform coefficient information by taking the difference between the orthogonal transform coefficient information and the generated approximate orthogonal transform coefficient information, and the residual orthogonal transform coefficient information as predetermined By performing entropy coding on the quantizing means for quantizing based on the quantization parameter of and the quantized residual orthogonal transform coefficient information output from the quantizing means, Entropy coding means for generating post-entropy-encoded information, and entropy-encoded information output from the entropy encoding means, by performing predetermined multiplexing according to a predetermined syntax structure to generate an encoded bitstream Multiplexing means to be generated, inverse quantization means for inversely quantizing residual orthogonal transform coefficient information after quantization output from the quantization means based on a predetermined quantization parameter, and first boundary condition correction information Is not valid, the video signal between the rectangular area to be decoded and the adjacent rectangular area based on the residual orthogonal transform coefficient information after the inverse quantization by the inverse quantization means Analyze discontinuity, determine whether correction of boundary conditions is necessary based on at least one predetermined determination condition, and specify whether correction of boundary conditions is necessary Boundary condition determination means for generating second boundary condition correction information for generating the boundary condition determination information using the generated second boundary condition correction information or effective first boundary condition correction information as boundary condition determination information; Decoding target that satisfies the Poisson's equation with the gradient of the video signal obtained from the residual orthogonal transform coefficient information after dequantization at the boundary of the adjacent rectangular region as the boundary condition For the estimated signal in the rectangular region, the approximate orthogonal transform coefficient information obtained by orthogonal transform is generated using the residual orthogonal transform coefficient information after the inverse quantization generated by the inverse quantization means, Inverse multiple harmonic local orthogonal transform means for decoding orthogonal transform coefficient information by combining the residual orthogonal transform coefficient information after inverse quantization and the generated approximate orthogonal transform coefficient information, and inverse multiple harmonic local orthogonal transform means Generated by By inverse orthogonal transformation on the orthogonal transformation coefficient information, and having an inverse orthogonal transform means for generating a local decoded image signal.

  In the present invention, instead of performing intra prediction in normal intra coding, orthogonal transform is performed in units of a rectangular area having a predetermined number of pixels with respect to a moving image signal of one screen area which is a coding target of intra coding. In addition, by analyzing the discontinuity between the rectangular area to be encoded and its adjacent and surrounding rectangular areas, it is determined whether the boundary condition needs to be corrected, and the correlation between adjacent rectangular areas When the boundary condition correction is unnecessary and the encoding target rectangular area is subjected to predetermined multiple harmonic local orthogonal transformation based on the boundary condition of the adjacent rectangular area, a residual orthogonal transformation coefficient is generated, As a result of the analysis of the boundary condition, when the correlation between adjacent rectangular regions is low and the boundary condition needs to be corrected, this conversion process is performed after correcting the boundary condition used in the multiple harmonic local orthogonal transformation.

  In order to achieve the above object, according to the moving image encoding apparatus of the fourth invention, in the third invention, the orthogonal transform means performs orthogonal transform using a rectangular region as a processing unit, and generates the generated orthogonal transform coefficient information. This is a means for outputting the processing unit of the rectangular area, and the boundary condition analyzing means, when analyzing the discontinuity existing in the moving image signal to be encoded, Analyzing a discontinuity between adjacent rectangular regions that have undergone orthogonal transformation and determining whether or not the boundary condition needs to be corrected based on at least one predetermined determination condition; Means for generating first boundary condition correction information for specifying whether or not correction is necessary, and the quantization means quantizes the residual orthogonal transform coefficient information based on a predetermined quantization parameter And after the generated quantization The inverse quantization means performs inverse quantization on the quantized residual orthogonal transformation coefficient information output from the quantization means based on a predetermined quantization parameter. And the generated residual orthogonal transform coefficient information after dequantization is output in units of processing in the rectangular area. The boundary condition determination unit is configured to perform inverse quantization when the first boundary condition correction information is not valid. Analysis of the discontinuity of the video signal between the rectangular area to be decoded and the adjacent rectangular area based on the residual orthogonal transform coefficient information after dequantization generated by the quantization means And determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition, and determining whether the boundary condition needs to be corrected. Generated information and the generated second boundary A means for the boundary condition determining information of the first boundary condition correction information is the correction information or effective, and performing coding as a processing unit a predetermined rectangular region.

  In the present invention, when encoding processing is performed in units of rectangular processing, when analyzing the boundary condition between the rectangular area to be encoded and its adjacent and surrounding rectangular areas, the rectangle after orthogonal transformation is used. The boundary condition is analyzed only between the region and the other rectangular region is interpreted as discontinuous.

  In order to achieve the above object, the moving picture coding apparatus according to the fifth invention is a coding object in which the multiple harmonic local orthogonal transform means satisfies the Poisson equation in the third and fourth inventions. When generating approximate orthogonal transform coefficient information obtained by orthogonal transform of an estimated signal in a rectangular area using the orthogonal transform coefficient information after orthogonal transform, the orthogonal transform coefficient information after orthogonal transform The orthogonal transform coefficient information after quantization is estimated based on the quantization parameter used for quantization, and the approximate orthogonal transform coefficient information is generated based on the estimation result.

  In this invention, the signal component in the encoding target rectangular region is subjected to multiple harmonic local orthogonal transformation, and the signal component is approximated by estimation based on the boundary condition with the adjacent rectangular region, thereby performing residual orthogonal transformation. Since the coefficient is generated, as compared with the case of selecting the closest one to the encoding target rectangular area from the prediction rectangular areas generated by the predetermined direction and the predetermined weight as in the conventional intra prediction, A residual orthogonal transform coefficient can be generated based on a more complicated estimation signal.

  In order to achieve the above object, according to a sixth aspect of the present invention, there is provided a moving picture encoding program for causing a computer to execute each of the constituent means of any one of the first to fifth aspects of the invention.

In order to achieve the above object, a moving picture decoding apparatus according to a seventh aspect of the invention is an information processing apparatus that operates according to the moving picture encoding apparatus of the first aspect of the invention or the moving picture encoding program of the sixth aspect of the invention. Is obtained by acquiring a coded bit stream generated by the above-described method from a predetermined storage medium or a predetermined transmission path, performing a decoding operation on the acquired coded bit stream, and outputting a decoded moving image signal. A device,
Demultiplexing means for demultiplexing encoded information on the encoded bitstream based on a predetermined syntax structure, and predetermined entropy decoding for the information after demultiplexing acquired from the demultiplexing means Entropy decoding means for generating post-quantization information, motion vector information, and parameter information necessary for configuring a predetermined syntax structure, and post-quantization information acquired from the entropy decoding means By performing inverse quantization, inverse quantization means for generating residual DCT coefficient information as residual orthogonal transform coefficient information after inverse quantization and a DCT base as an orthogonal transform base are used for decoding. Decoding that satisfies the Poisson's equation with the gradient of the video signal obtained from the orthogonal transformation coefficient information at the boundary between the rectangular region and another rectangular region as a boundary condition For the estimated signal of the rectangular region of interest, the orthogonal transformation to approximate DCT coefficient information, after generating by using the residual DCT coefficient information after the inverse quantization, obtained by the inverse quantization residual the difference DCT coefficient information generated approximate DCT coefficient information and the decoded DCT coefficient information obtained synthesized with reverse polyharmonic local orthogonal transform means for generating a decoded DCT coefficient information by performing the ,, demultiplexing conditioning local orthogonal transform means An inverse orthogonal transform unit that generates a decoded differential image signal for one screen by performing inverse orthogonal transform on the predicted image signal generated by a predetermined motion compensation using motion vector information acquired from the entropy decoding unit, Or it has the image decoding means which synthesize | combines the prediction image signal produced | generated by intra prediction, and a decoding difference image signal, and produces | generates a decoding moving image signal, It is characterized by the above-mentioned.

  In this invention, when encoding a moving image with motion estimation / compensation prediction, orthogonal transform coefficient information is decoded from an encoded bit rate encoded by applying a multiple harmonic local orthogonal transform method, The moving image signal can be decoded by performing orthogonal transformation, decoding the difference image signal, and synthesizing it with the predicted image signal. Here, the orthogonal transform coefficient information may be DCT coefficient information.

In order to achieve the above object, a moving picture decoding apparatus according to an eighth aspect of the present invention is a moving picture encoding apparatus according to the second aspect of the invention or an information processing apparatus that operates according to the moving picture encoding program of the sixth aspect of the invention Is obtained by acquiring a coded bit stream generated by the above-described method from a predetermined storage medium or a predetermined transmission path, performing a decoding operation on the acquired coded bit stream, and outputting a decoded moving image signal. A device,
Demultiplexing means for demultiplexing encoded information on the encoded bitstream based on a predetermined syntax structure, and predetermined entropy decoding for the information after demultiplexing acquired from the demultiplexing means Entropy decoding means for generating post-quantization information, motion vector information and parameter information necessary for constructing a predetermined syntax structure, and extracting encoded and incorporated boundary condition correction information Inverse quantization means for generating residual orthogonal transform coefficient information after inverse quantization by performing inverse quantization on the quantized information obtained from entropy decoding means, and obtained from entropy decoding means If the boundary condition correction information is not valid, it is a decoding target based on the residual orthogonal transform coefficient information after quantization obtained from the inverse quantization means. Analyzing the discontinuity of the moving image signal between the shape region and the adjacent surrounding rectangular region, and determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition; Generate boundary condition correction information for specifying whether or not boundary condition correction is necessary, and determine the boundary condition correction information obtained from the generated boundary condition correction information or entropy decoding means. Obtain boundary condition determination means as information, boundary condition determination information and residual orthogonal transform coefficient information after inverse quantization, and orthogonality at the boundary between the rectangular area adjacent to the rectangular area to be decoded For the estimated signal in the rectangular area to be decoded that satisfies the Poisson equation with the gradient of the transform coefficient information as the boundary condition, the approximate orthogonal transform coefficient information obtained by orthogonal transform Inverse multiple harmony that generates decoded orthogonal transform coefficient information by combining residual orthogonal transform coefficient information after dequantization and generated approximate orthogonal transform coefficient information after generating using difference orthogonal transform coefficient information Local orthogonal transform means, inverse orthogonal transform means for generating a decoded differential image signal for one screen by performing inverse orthogonal transform on the decoded orthogonal transform coefficient information acquired from the demultiplex harmonic local orthogonal transform means, and entropy decoding An image that generates a decoded moving image signal by synthesizing a prediction image signal generated by predetermined motion compensation using motion vector information acquired from the converting means or a prediction image signal generated by intra prediction and a decoded difference image signal And a decoding means.

  In the present invention, when the boundary condition correction information acquired from the entropy decoding unit is not valid, the boundary condition correction information is a decoding target based on the quantized residual orthogonal transform coefficient information acquired from the inverse quantization unit. Analyzing the discontinuity of the moving image signal between the rectangular area and the adjacent surrounding rectangular area, and determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition; Generate boundary condition correction information for specifying whether or not boundary condition correction is necessary, and determine the boundary condition correction information obtained from the generated boundary condition correction information or entropy decoding means. By using the information, it is possible to correctly reproduce and decode the discontinuity of signals between adjacent rectangular areas. Here, the decoded orthogonal transform coefficient information may be decoded DCT coefficient information, and the decoded approximate orthogonal transform coefficient information may be decoded approximate DCT coefficient information.

In order to achieve the above object, the moving picture decoding apparatus according to the ninth aspect of the present invention is the moving picture encoding apparatus according to the third, fourth and fifth aspects of the invention, or the moving picture encoding program of the sixth aspect of the invention. The encoded bit stream generated by the information processing apparatus operating according to the above is acquired from a predetermined storage medium or a predetermined transmission path, and the decoded video signal is obtained by performing a decoding operation on the acquired encoded bit stream. An output video decoding device, comprising:
Demultiplexing means for demultiplexing encoded information on the encoded bitstream based on a predetermined syntax structure, and predetermined entropy decoding for the information after demultiplexing acquired from the demultiplexing means Entropy decoding means for generating post-quantization information and parameter information necessary for constructing a predetermined syntax structure, and extracting encoded and incorporated boundary condition correction information, and entropy decoding Inverse quantization means for generating residual orthogonal transform coefficient information after inverse quantization by performing inverse quantization on the quantized information obtained from the means, and boundary condition correction information obtained from the entropy decoding means If is not valid, based on the post-quantization residual orthogonal transform coefficient information obtained from the inverse quantization means, the surrounding area adjacent to the rectangular area to be decoded The discontinuity of the moving image signal between the rectangular area is analyzed, it is determined whether the boundary condition needs to be corrected based on at least one predetermined determination condition, and the boundary condition needs to be corrected. Boundary condition determination means for generating boundary condition correction information for specifying whether or not there is, and using the generated boundary condition correction information or effective boundary condition correction information acquired from the entropy decoding means as boundary condition determination information And the boundary condition determination information and the residual orthogonal transform coefficient information after inverse quantization, and the moving image obtained from the orthogonal transform coefficient information at the boundary between the rectangular area to be decoded and another adjacent rectangular area For the estimated signal in the rectangular region to be decoded that satisfies the Poisson equation with the signal slope as the boundary condition, the approximate orthogonal transform coefficient information obtained by orthogonal transform is the remaining after the inverse quantization. After generating using orthogonal transform coefficient information, demultiplexed harmonic transform local information that generates decoded orthogonal transform coefficient information by combining the residual orthogonal transform coefficient information after dequantization and the generated approximate orthogonal transform coefficient information An orthogonal transform unit and an inverse orthogonal transform unit that generates a decoded image signal for one screen by performing inverse orthogonal transform on the decoded orthogonal transform coefficient information acquired from the demultiplex harmonic local orthogonal transform unit. And

  A moving picture decoding program according to a tenth aspect of the present invention is a moving picture decoding which causes a computer to execute each component of the moving picture decoding apparatus according to any one of the seventh to ninth aspects of the moving picture decoding apparatus. It is characterized by comprising a computer program.

  According to the present invention, the gradient of the moving image signal obtained from the orthogonal transform coefficient information at the boundary between the rectangular region to be encoded and another rectangular region adjacent to the rectangular region to be encoded is determined using the method of multiple harmonic local orthogonal transformation. Using the orthogonal transform coefficient information obtained from the orthogonal transform means, the approximate orthogonal transform coefficient information obtained by orthogonal transform for the estimated signal in the rectangular region to be encoded that satisfies the Poisson equation Generate. The above estimated signal is an approximate waveform that cancels the discontinuity of the boundary part of the rectangular area to be encoded and improves the smoothness of the waveform at the boundary part. After generating an approximate orthogonal transform coefficient, the residual orthogonal transform coefficient information is generated by taking the difference between the orthogonal transform coefficient information and the generated approximate orthogonal transform coefficient information. Generation of unnecessary high frequency components of the orthogonal transform coefficient can be suppressed.

  Further, according to the present invention, after the residual orthogonal transform coefficient information in which unnecessary high frequency components are suppressed is quantized based on a predetermined quantization parameter as an encoding target instead of a conventional orthogonal transform coefficient. Since the encoded bitstream is generated according to a predetermined syntax structure by performing entropy encoding on the quantized residual orthogonal transform coefficient information, the moving image coding is performed with a smaller amount of code than in the past. Thus, encoding efficiency can be improved.

  Further, according to the present invention, the discontinuity of the moving image signal between the rectangular area to be encoded and the adjacent surrounding rectangular area is analyzed, and the boundary is determined based on at least one predetermined determination condition. Since it is determined whether or not correction of conditions is necessary and whether or not correction of boundary conditions is necessary, signal discontinuity between adjacent rectangular regions can be correctly reproduced.

  In other words, in the present invention, the encoding target is the case where the continuity of the original signal is not maintained between the rectangular area to be encoded and the adjacent surrounding rectangular areas. Comparing the boundary condition of the rectangular area with the boundary condition of the surrounding rectangular area, for example, the boundary condition of the rectangular area that is the encoding target is the one with the smaller gradient of the image signal that is the boundary condition of the rectangular area If the slope is compared and the slope exceeds a predetermined threshold, the slope is set to zero, and this is adopted as a boundary condition, so that the multiple harmonic local cosine transform method is applied. It is possible to correctly reproduce the discontinuity between the rectangular areas, and to improve the problem that this influence is propagated as an error by motion compensated prediction of moving picture decoding.

  Furthermore, according to the present invention, by using DCT as the transform base of the multiple harmonic local orthogonal transform, the conventional moving image is maintained while maintaining compatibility with the orthogonal transform generally used in the conventional moving image coding method. Encoding efficiency in encoding can be improved.

  Also, the encoded bitstream generated by the moving image encoding apparatus and the moving image encoding program of the present invention is acquired from a predetermined transmission path or recording medium, and the moving image decoding apparatus of the present invention or the moving image of the present invention is acquired. By decoding by a computer using a decoding program, it is possible to efficiently transmit, receive, and play back an encoded bit stream having a smaller code amount than conventional ones.

  Further, according to the present invention, instead of performing intra prediction in normal intra coding, a rectangular area unit having a predetermined number of pixels with respect to a moving picture signal of one screen area that is a coding target of intra coding. In addition to performing orthogonal transformation at the same time, analysis of discontinuity between the rectangular area to be encoded and its neighboring areas and surrounding rectangular areas determines whether or not boundary conditions need to be corrected, and adjacent rectangles are determined. If the correlation between regions is high and correction of boundary conditions is not required, a predetermined multiharmonic local orthogonal transform is performed on the rectangular region to be encoded based on the boundary conditions of adjacent rectangular regions, and a residual orthogonal transform coefficient is obtained. If the correlation between the adjacent rectangular areas is low and the boundary condition needs to be corrected as a result of the boundary condition analysis, the boundary condition used in the multiple harmonic local orthogonal transform must be corrected before this conversion process is performed. By By processing while taking into account discontinuities between the shape regions, the bias of the signal component of the generated residual orthogonal transform coefficient is made larger than the intra prediction used in conventional intra coding, and sufficient code Efficiency can be obtained.

  Further, according to the present invention, when encoding processing is performed in units of rectangular processing, orthogonality is analyzed when analyzing boundary conditions between a rectangular region to be encoded and its adjacent and surrounding rectangular regions. By analyzing the boundary condition only with the converted rectangular area and interpreting it as discontinuous with the other rectangular areas, the processing of the present invention is reduced, and faster encoding and decoding are performed. Can be realized.

  Furthermore, according to the present invention, the signal component in the encoding target rectangular region is subjected to multiple harmonic local orthogonal transformation, and the signal component is approximated by estimation based on a boundary condition with an adjacent rectangular region. Since the residual orthogonal transform coefficient is generated, the closest one to the encoding target rectangular area is selected from the predicted rectangular areas generated by the predetermined direction and the predetermined weighting as in the conventional intra prediction. Since the residual orthogonal transform coefficient can be generated based on a more complicated estimation signal than in the case, higher encoding efficiency can be obtained.

  Next, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a moving picture encoding apparatus according to the present invention. In the figure, the same components as those in FIG. 29 are denoted by the same reference numerals. As shown in FIG. 1, the present embodiment is provided with a multiple harmonic local orthogonal transformer 121 between the orthogonal transformer 103 and the quantizer 104, compared with the conventional coding apparatus shown in FIG. The difference is that an inverse multiple harmonic local orthogonal transformer 122 is provided between the inverse quantizer 105 and the inverse orthogonal transformer 106, and the output code of the entropy encoder 123 is transmitted through the multiplexer 124 for output transmission. Is input to the device 125.

  The switch 113 has a function of switching predicted image frames generated by the intra predictor 110 and the motion compensator (MC unit) 112 according to the encoding mode of the picture to be encoded. When the picture coding mode is intra coding, the switch 113 is switched to the terminal a. If the picture coding mode is not intra coding, the switch 113 is switched to the terminal b.

  The orthogonal transformer 103 acquires a difference image frame between the moving image frame obtained as the input code 101 and the prediction image frame obtained from the intra predictor or the MC unit 112, and performs a predetermined process on the acquired difference image frame. It has a function of performing orthogonal transformation. Here, the predetermined orthogonal transform is assumed to be DCT. The orthogonal transformer 103 has a function of notifying the multiharmonic local orthogonal transformer 121 described later of coefficient information after orthogonal transformation. Here, the coefficient information after orthogonal transform is DCT coefficient information.

  The multi-harmonic local orthogonal transformer 121 has a function of obtaining coefficient information (here, DCT coefficient information) after orthogonal transformation from the orthogonal transformer 103, and analyzes the Poisson equation based on the boundary condition of each block. It has a function of obtaining approximate orthogonal transform coefficient information (here, approximate DCT coefficient information) in the block by solving the above. Further, the multiple harmonic local orthogonal transformer 121 generates residual orthogonal transform coefficient information (here, residual DCT coefficient information) from the difference between the acquired DCT coefficient information and the obtained approximate DCT coefficient information, It has a function of notifying the quantizer 104 of residual DCT coefficient information.

  In addition, the multi-harmonic local orthogonal transformer 121 obtains approximate DCT coefficient information obtained by orthogonal transform (in this case, DCT) of an estimated signal in a predetermined rectangular region that is an encoding target that satisfies the Poisson equation. When generating using the DCT coefficient information after orthogonal transformation, a quantization parameter used for quantization is obtained for the DCT coefficient information after orthogonal transformation, and the quantized DCT is obtained based on the quantization parameter. It would be a better configuration if it had a function to estimate coefficient information and generate approximate DCT coefficient information based on this estimation result.

  This multi-harmonic local cosine transform method is a method described in Non-Patent Document 3 described above, and introduces the concept of the Poisson equation to obtain an estimated signal for the original signal in the block. By generating a residual signal that is a difference between the two, a DCT coefficient having a higher frequency can be quickly converged when DCT is introduced, and as a result, encoding efficiency can be improved.

  That is, as described above, in the multiple harmonic local cosine transform method, the gradient 302 of the image signal at both ends of the original signal 301 corresponding to the block boundaries 304 and 305 with respect to the original signal 301 shown in FIG. 303, and the Poisson equation concept is introduced using the gradients 302 and 303 as boundary conditions to generate an estimated signal 306 as shown in FIG. 31 (b). Then, as shown in FIG. A residual signal 307 between the signal 301 and the estimated signal 306 is obtained.

  When DCT is performed by orthogonally transforming the residual signal 307, even connection is performed with one of the block boundaries (here, the right side of the signal) of the residual signal 307 shown in FIG. Then, by using a residual waveform periodic waveform signal as shown in FIG. 5B, it is possible to express it by a DCT series. The periodic waveform signal of the residual signal is smoother than the case where the DCT is performed on the original signal in the dotted line portions 501 to 503 in FIG. It is possible to prevent mixing of unnecessary signal components not included in.

  Therefore, when the spectrum is expressed by a DCT series, it becomes as shown by I in FIG. 33 (c), which is indicated by a dotted line portion 505 in FIG. 33 (c) as compared with II in the case of performing normal DCT. Even in such a high frequency component region, the DCT coefficients are sufficiently converged, and by performing predetermined quantization and entropy coding on the DCT coefficients using the multiple harmonic local cosine transform method, normal DCT coding is performed. It is possible to suppress the amount of code generated compared to the case where it is performed, leading to improvement in coding efficiency.

  In this embodiment, since it is relatively easy to extend from the conventional coding apparatus, DCT is performed on the original signal in the block and the estimated signal from the boundary condition as described above as an example. However, the present invention is not particularly limited to this configuration, and the residual signal, which is the difference between the original signal and the estimated signal, is employed. However, the configuration may be such that DCT is performed.

  Referring back to FIG. 1 again, the quantizer 104 acquires residual DCT coefficient information from the multiharmonic local orthogonal transformer 121 and performs predetermined quantization on the acquired residual DCT coefficient information. Thus, it has a function of generating post-quantization information which is residual DCT coefficient information after quantization. The quantizer 104 has a function of supplying the generated post-quantization information to the inverse quantizer 105 and the entropy encoder 123, respectively. The inverse quantizer 105 generates predetermined post-quantization information that is residual DCT coefficient information after inverse quantization by performing predetermined inverse quantization on the post-quantization information acquired from the quantizer 104. Thus, it has a function of supplying it to the inverse multiple harmonic local orthogonal transformer 122.

  The inverse multiple harmonic local orthogonal transformer 122 acquires post-inverse quantization information from the inverse quantizer 105, and performs predetermined inverse multiple harmonic local orthogonal transform on the acquired post-quantization information, Decoded DCT coefficient information obtained by decoding from residual DCT coefficient information after inverse quantization is generated. That is, the demultiplex harmonic local orthogonal transformer 122 performs residual orthogonal transform coefficient information (here, residual DCT coefficient information) after inverse quantization at the boundary between the block to be decoded and another adjacent block. ) Approximate orthogonal transform coefficient information (here, approximate DCT coefficient information) obtained by orthogonal transform with respect to the estimation signal in the block to be decoded so as to satisfy the Poisson equation having the slope of Residual orthogonal transform coefficient information (here, residual DCT coefficient information) (in this case, residual DCT coefficient information) after inverse quantization generated by the inverse quantizer 105 and generated after inverse quantization. Then, decoding orthogonal transform coefficient information (here, DCT coefficient information) is generated by combining the residual DCT coefficient information) and the generated approximate orthogonal transform coefficient information (here, approximate DCT coefficient information). That. The inverse multiple harmonic local orthogonal transformer 122 supplies the generated decoded DCT coefficient information to the inverse orthogonal transformer 106.

  The inverse orthogonal transformer 106 has a function of acquiring decoded DCT coefficient information from the inverse multiple harmonic local orthogonal transformer 122 and generating a decoded difference image frame by performing inverse orthogonal transformation on the acquired decoded DCT coefficient information. Have. Further, the inverse orthogonal transformer 106 has a function of supplying the generated decoded difference image frame to the adder 107. The adder 107 has a function of adding the output signal of the intra predictor 110 or the MC unit 112 selected by the switch 113 to generate a decoded image frame and supplying the decoded image frame to the intra predictor 110 and the deblocking filter 108.

  The ME (motion estimator) 111 obtains the moving image frame that is the current encoding target obtained as the input code 101 and the predetermined reference image frame stored in the frame memory 109, and the obtained moving image. It has a function of performing predetermined ME between the frame and the reference image frame to generate motion vector information. The ME device 111 has a function of notifying the generated motion vector information to the MC device 112 and the entropy encoder 123.

  The MC (motion compensation) unit 112 has a function of acquiring motion vector information from the ME unit 111. In addition, the ME device 111 has a function of acquiring the reference image frame used when the motion vector information is generated from the frame memory 109. From the acquired motion vector information and reference image frame, a predicted image frame is generated and output to the terminal b side of the switch 113. The frame memory 109 has a function of acquiring and accumulating the deblocked image frame as a reference image frame from the deblocking filter 108. Further, it has a function of notifying the ME device 111, the MC device 112, and the intra predictor 110 of reference image frames as necessary. Here, it is desirable that the frame memory 109 can store at least one reference image frame and can notify a necessary reference image frame in response to a request from each unit.

  The entropy encoder 123 acquires at least the post-quantization information from the quantizer 104 and the motion vector information from the ME unit 111, and configures the acquired post-quantization information and motion vector information, and other predetermined syntax structures Predetermined entropy encoding is performed on the parameter information necessary for the purpose, and post-entropy encoded information is generated. The entropy encoder 123 has a function of notifying the multiplexer 124 of the generated entropy encoding information.

  The multiplexer 124 acquires the entropy-encoded information from the entropy encoder 123, and multiplexes the encoded information on the acquired entropy-encoded information based on a predetermined syntax structure. It has a function of generating an encoded bit stream. Here, the parameter information necessary for constructing a predetermined syntax structure includes macroblock information for specifying various states of the macroblock, quantization parameter information used when performing quantization and inverse quantization, intra More preferably, intra prediction mode information for specifying a prediction method, frame order information for specifying the reference order of reference image frames, and the like are included. The multiplexer 124 has a function of notifying the output transmitter 125 of the generated encoded bit stream.

  The output transmitter 125 acquires the encoded bit stream from the multiplexer 124 and outputs the acquired encoded bit stream as an output code to a predetermined transmission path or a predetermined recording medium. It has a function of generating packet information by performing packetization processing so as to have a configuration having a payload. Here, the packet information treats the packet header and payload as one unit. Usually, the obtained encoded bit stream is divided into smaller units, stored in a payload, and packetized by adding a predetermined packet header. The output transmitter 125 has a function of outputting each generated packet information as an output code according to the state of a transmission path or a recording medium as a predetermined output destination.

  The intra predictor 110 obtains a decoded image frame synthesized from the decoded difference image frame generated by the inverse orthogonal transformer 106 and the predicted image frame generated by the intra predictor 110 or the MC unit 112 selected by the switch 113, and It has a function of performing intra prediction using the decoded image in a region where encoding has been completed for prediction on a block that has not been encoded in the acquired decoded image frame. At this time, it is further preferable to have a function of generating intra prediction mode information for specifying an intra prediction method and notifying the entropy encoder 123 of the information.

  In addition, when performing intra prediction processing using the lower layer decoded image information such as prediction processing for an intra block or scalable encoding, a moving image frame that is the current encoding target from the frame memory 109 May have a function of acquiring a reference image frame necessary for encoding. In the case of scalable coding, the reference image frame may be a decoded image frame of a layer whose coding has been completed. The intra predictor 110 has a function of outputting a predicted image frame obtained by intra prediction to the terminal a side of the switch 113.

  The deblocking filter 108 has a function of acquiring a decoded image frame synthesized from the decoded differential image frame generated by the inverse orthogonal transformer 106 and the predicted image frame generated by the intra predictor 110 or the MC unit 112 selected by the switch 113. And it has a function which produces | generates the decoded image frame after a deblocking process by performing predetermined | prescribed deblocking processing with respect to the acquired decoded image frame. Here, the predetermined deblocking process is performed according to the encoding mode of the frame or block currently being encoded, the state of the quantization parameter, the state of the image information in the surrounding encoded blocks, and the like. It is preferable that the block processing can be turned on / off and the strength of the filter processing can be switched.

  In addition, when the transform processing is correctly performed in the multi-harmonic local orthogonal transformer 121, it is generally determined that the deblocking process is not necessary, and the function is simply performed as an all-band pass filter without performing the deblocking process. It should be configured. The deblocking filter 108 has a function of notifying the frame memory 109 so that the generated decoded image frame after the deblocking process can be used as a reference image frame.

  Next, the operation of the embodiment of the encoding apparatus of the present invention shown in FIG. 1 will be described with reference to the flowchart shown in FIG. First, an image frame to be encoded is prepared as the input code 101 (step S11), and it is determined whether or not intra encoding is performed (step S12). In the case of intra coding, intra prediction processing is performed (step S13). The predicted image frame obtained by the intra predictor 110 is output to the terminal a side of the switch 113. Further, the switch 113 is switched to the terminal a side. Thereafter, the process proceeds to step S16.

  On the other hand, if it is determined in step S12 that the encoding is not intra coding, the ME device 111 acquires a moving image frame that is the current encoding target and a predetermined reference image frame stored in the frame memory 109. Then, motion estimation is performed (step S14). The ME device 111 notifies the MC device 112 of the motion vector information generated by the motion estimation process. In addition, the entropy encoder 123 is notified in order to entropy encode the motion vector information.

  The MC unit 112 acquires the motion vector information generated by the ME unit 111 through the motion estimation process. Further, the reference image frame used when the ME device 111 generates the motion vector information is acquired from the frame memory 109. The MC unit 112 generates a predicted image frame from the acquired motion vector information and the reference image frame (step S15), and outputs the predicted image frame to the terminal b side of the switch 113. Further, the switch 113 is switched to the terminal b side. Thereafter, the process proceeds to step S16.

  The orthogonal transformer 103 includes a moving image frame obtained as the input code 101, a predicted image frame obtained from the intra predictor 110 connected to the terminal a of the switch 113 or the MC unit 112 connected to the terminal b of the switch 113, and The difference image frame is acquired. A predetermined orthogonal transformation is performed on the acquired difference image frame (step S16). The orthogonal transformer 103 notifies the multiple harmonic local orthogonal transformer 121 of the coefficient information after the orthogonal transformation. Here, the coefficient information after orthogonal transform is DCT coefficient information.

  The multiharmonic local orthogonal transformer 121 acquires coefficient information after orthogonal transform, here DCT coefficient information, from the orthogonal transformer 103, and performs multiple harmonic local orthogonal transform processing (step S17). The generated residual DCT coefficient information is notified to the quantizer 104 described later. The quantizer 104 obtains residual DCT coefficient information from the multiharmonic local orthogonal transformer 121 and performs predetermined quantization on the residual DCT coefficient information, thereby performing residual DCT coefficient information after quantization. The post-quantization information is generated (step S18). The generated post-quantization information is supplied to the inverse quantizer 105 and the entropy encoder 123.

  The inverse quantizer 105 obtains post-quantization information from the quantizer 104 and performs predetermined inverse quantization on the post-quantization information, thereby obtaining residual DCT coefficient information after inverse quantization. Information after inverse quantization is generated (step S19). The generated information after inverse quantization is supplied to the inverse multiple harmonic local orthogonal transformer 122.

  The inverse multiple harmonic local orthogonal transformer 122 obtains the information after inverse quantization from the inverse quantizer 105 and performs a predetermined inverse multiple harmonic local orthogonal transform process on the information after inverse quantization, thereby performing inverse processing. Decoded DCT coefficient information obtained by decoding from the quantized residual DCT coefficient information is generated (step S20). The generated decoded DCT coefficient information is supplied to the inverse orthogonal transformer 106.

  The inverse orthogonal transformer 106 obtains the decoded DCT coefficient information from the inverse multiple harmonic local orthogonal transformer 122, and generates a decoded difference image frame by performing inverse orthogonal transformation on the decoded DCT coefficient information (step S21). . The generated decoded difference image frame is supplied to the intra predictor 110 and the deblocking filter 108, respectively.

  The deblocking filter 108 is a decoded image obtained by combining the decoded difference image frame generated by the inverse orthogonal transformer 106 and the predicted image frame generated by the intra predictor 110 or the MC unit 112 selected by the switch 113 by the adder 107. A frame is acquired, and a predetermined deblocking process is performed on the acquired decoded image frame as necessary, thereby generating a decoded image frame after the deblocking process (step S22). However, it is often determined that there is no need for the normal deblocking process, and it often functions simply as an all-band pass filter without performing the deblocking process. The deblocking filter 108 stores the generated decoded image frame after the deblocking process in the frame memory 109 so that it can be used as a reference image frame.

  The entropy encoder 123 acquires at least the post-quantization information from the quantizer 104 and the motion vector information from the ME unit 111, and performs predetermined entropy encoding on the post-quantization information and the motion vector information. (Step S23). The multiplexer 124 multiplexes the encoded information on the post-entropy encoded information acquired from the entropy encoder 123 based on a predetermined syntax structure, thereby generating an encoded bit stream (step S24). ). The generated encoded bit stream is output as an output code to a predetermined transmission path or recording medium by the output transmitter 125.

  By sequentially performing the encoding process as described above on the moving image frame to be encoded, the encoding apparatus according to the present embodiment generates an encoded bit stream with higher encoding efficiency than the conventional encoding apparatus. It can be obtained as an output code.

  Next, an embodiment of a decoding apparatus according to the present invention will be described. FIG. 3 shows a block diagram of an embodiment of a decoding apparatus according to the present invention. In the figure, the same components as those in FIG. 30 are denoted by the same reference numerals. In FIG. 3, an input receiver 222 acquires packet information as an input code 221 from a predetermined transmission path and a predetermined recording medium, and as an encoded bit stream generated by the encoding apparatus of the present invention from the acquired packet information. It has a function to reconfigure. The input receiver 222 also supplies the reconstructed encoded bitstream to the demultiplexer 223.

  The demultiplexer 223 acquires the encoded bit stream from the input receiver 222, performs demultiplexing of the encoded information on the acquired encoded bit stream based on a predetermined syntax structure, and demultiplexes It has a function of notifying the entropy decoder 224 of later information. The entropy decoder 224 acquires the information after demultiplexing from the demultiplexer 223, performs predetermined entropy decoding on the acquired information after demultiplexing, and performs post-quantization information and motion vector It has a function of generating information and parameter information necessary for configuring other predetermined syntax structures.

  Here, the parameter information necessary for constructing a predetermined syntax structure includes macroblock information for specifying various states of the macroblock, quantization parameter information used when performing quantization and inverse quantization, intra More preferably, intra prediction mode information for specifying a prediction method, frame order information for specifying the reference order of reference image frames, and the like are included. The entropy decoder 224 has a function of notifying at least the generated post-quantization information to the inverse quantizer 225 and also has a function of notifying the MC unit 206 of at least the generated motion vector information.

  The inverse quantizer 225 acquires post-quantization information from the entropy decoder 224, and performs predetermined inverse quantization on the acquired post-quantization information, thereby performing a residual DCT coefficient after inverse quantization. It has a function of generating post-inverse quantization information that is information. The inverse quantizer 225 supplies the generated post-quantization information to the inverse multiple harmonic local orthogonal transformer 226.

  The inverse multiple harmonic local orthogonal transformer 226 acquires information after inverse quantization from the inverse quantizer 225, and performs predetermined inverse multiple harmonic local orthogonal transform on the acquired inverse quantization information, Decoded DCT coefficient information obtained by decoding from residual DCT coefficient information after inverse quantization is generated. The inverse multiple harmonic local orthogonal transformer 226 supplies the generated decoded DCT coefficient information to the inverse orthogonal transformer 204.

  The inverse orthogonal transformer 204 has a function of acquiring the decoded DCT coefficient information from the inverse multiple harmonic local orthogonal transformer 226, and generating a decoded difference image frame by performing inverse orthogonal transformation on the acquired decoded DCT coefficient information. Have. The inverse orthogonal transformer 204 supplies the generated decoded difference image frame to the intra predictor 208 and the deblocking filter 209. Here, a predicted image frame generated by the intra predictor 208 or the MC unit 206 selected by the switch 207 is synthesized by the adder 205 with respect to the decoded difference image frame generated by the inverse orthogonal transformer 204, and the decoded image frame is decoded. Is preferably supplied to the intra predictor 208 and the deblocking filter 209. In the present embodiment, the description proceeds on the assumption that the decoded image frame generated by the adder 205 is supplied to the intra predictor 208 and the deblocking filter 209.

  The MC (motion compensation) unit 206 acquires the motion vector information from the entropy decoder 224 and decodes the reference image frame to be equivalent to the reference image frame used when the acquired motion vector information is generated. And a function of acquiring the stored reference image frame from the frame memory 210. The MC unit 206 has a function of generating a predicted image frame from the acquired motion vector information and reference image frame, and outputting the predicted image frame to the terminal b side of the switch 207.

  The switch 207 has a function of switching the predicted image frames generated by the intra predictor 208 and the MC unit 206 according to the encoding mode of a picture obtained from the encoded bitstream. If the picture coding mode is intra coding, the switch 207 is switched to the terminal a. If the picture coding mode is not intra coding, the switch 207 is switched to the terminal b.

  The intra predictor 208 acquires and acquires a decoded image frame synthesized from the decoded difference image frame generated by the inverse orthogonal transformer 204 and the predicted image frame generated by the intra predictor 208 or the MC unit 206 selected by the switch 207. In the decoded image frame, the block has not been encoded yet, and has a function of performing intra prediction using the decoded image in the area where encoding is completed for prediction. At this time, in order to acquire intra prediction mode information for specifying an intra prediction method, the intra prediction mode information obtained by the multiplexing separation and entropy decoding processing is acquired from the encoded bitstream by the entropy decoder 224. It is better to have such a function.

  In addition, when performing intra prediction processing using intra-block prediction processing or lower-layer decoded image information such as scalable encoding, a moving image frame currently being decoded from the frame memory 210 May have a function of acquiring a reference image frame necessary for decoding. In the case of scalable coding, the reference image frame may be a decoded image frame of a hierarchy that has already been decoded. It has a function of outputting a predicted image frame obtained by intra prediction to the terminal a side of the switch 207.

  The deblocking filter 209 acquires a decoded image frame synthesized from the decoded differential image frame generated by the inverse orthogonal transformer 204 and the predicted image frame generated by the intra predictor 208 or the MC unit 206 selected by the switch 207. It has a function of generating a decoded image frame after the deblocking process by performing a predetermined deblocking process on the decoded image frame.

  Here, the predetermined deblocking process is performed according to the encoding mode of the frame or block currently being decoded, the state of the quantization parameter, the state of the image information in the surrounding decoded blocks, and the like. It is preferable that the block processing can be turned on / off and the strength of the filter processing can be switched. In addition, when the transform processing is correctly performed in the multi-harmonic local orthogonal transformer 121, it is generally determined that the deblocking process is not necessary, and the function is simply performed as an all-band pass filter without performing the deblocking process. It should be configured. The deblocking filter 209 has a function of notifying the frame memory 210 so that the generated decoded image frame after the deblocking process can be used as a reference image frame.

  The frame memory 210 has a function of acquiring and storing an image frame after deblocking from the deblocking filter 209 as a reference image frame. Further, it has a function of notifying a reference image frame to the MC unit 206 and the intra predictor 208 as necessary. Here, it is desirable that the frame memory 210 accumulates at least one reference image frame and can notify a necessary reference image frame in response to a request from each component. In addition, it is desirable to have a function of notifying an external output display device of an image frame necessary for display as an output code 211 according to the display timing of the decoded image frame.

  Next, the operation of the embodiment of the decoding apparatus of the present invention shown in FIG. 3 will be described with reference to the flowchart of FIG. First, the input receiver 222 sequentially acquires predetermined packet information from a predetermined transmission path or a predetermined recording medium as the input code 221 to reconstruct the encoded bit stream. Thereafter, the encoded bit stream is notified to the demultiplexer 223. The demultiplexer 223 acquires the encoded bit stream reconstructed by the input receiver 222, and performs demultiplexing of encoded information on the acquired encoded bit stream based on a predetermined syntax structure ( In step S31), the demultiplexed information is supplied to the entropy decoder 224.

  The entropy decoder 224 obtains information after demultiplexing and performs predetermined entropy decoding, so that it is necessary to construct post-quantization information and motion vector information, and other predetermined syntax structures. Parameter information and the like are generated (step S32). Thereafter, the entropy decoder 224 supplies at least post-quantization information to the inverse quantizer 225 and also supplies motion vector information to the MC unit 206.

  The inverse quantizer 225 obtains post-quantization information from the entropy decoder 224, and performs predetermined inverse quantization on the obtained post-quantization information, so that residual DCT coefficient information after inverse quantization is obtained. Is generated (step S33), and the generated post-quantization information is supplied to the inverse multiple harmonic local orthogonal transformer 226.

  Thereafter, it is determined whether or not intra decoding is performed (step S34). If it is determined that intra decoding is performed, intra prediction processing is performed (step S35). The predicted image frame obtained by the intra prediction by the intra predictor 208 is supplied to the terminal a side of the switch 207. Further, the switch 207 is switched to the terminal a side. Thereafter, the process proceeds to step S37.

  On the other hand, if it is determined in step S34 that the decoding is not intra decoding, the MC unit 206 acquires the motion vector information from the entropy decoder 224 and the reference image frame used when the motion vector information is generated. A reference image frame decoded and stored so as to constitute an equivalent reference image frame is acquired from the frame memory 210, and motion compensation is performed from the acquired motion vector information and the reference image frame (step S36). As a result, a prediction image frame is generated by the MC unit 206 and supplied to the terminal b side of the switch 207. Further, the switch 207 is switched to the terminal b side. Thereafter, the process proceeds to step S37.

  Next, the inverse multiple harmonic local orthogonal transformer 226 acquires post-inverse quantization information from the inverse quantizer 225, and performs a predetermined inverse multiple harmonic local orthogonal transform on the post-quantization information (step S37), thereby generating decoded DCT coefficient information obtained by decoding from the residual DCT coefficient information after inverse quantization, and supplying the generated decoded DCT coefficient information to the inverse orthogonal transformer 204.

  The inverse orthogonal transformer 204 acquires the decoded DCT coefficient information from the inverse multiple harmonic local orthogonal transformer 226, performs inverse orthogonal transformation on the decoded DCT coefficient information (step S38), and thereby generates a decoded difference image frame. Then, the generated decoded difference image frame is supplied to the intra predictor 208 and the deblocking filter 209.

  Here, as shown in FIG. 3, the adder 205 combines the decoded difference image frame generated by the inverse orthogonal transformer 204 and the predicted image frame generated by the intra predictor 208 or the MC unit 206 selected by the switch 207. It is desirable that the decoded image frame is synthesized and then supplied to the intra predictor 208 and the deblocking filter 209.

  The deblocking filter 209 adds an decoded image frame obtained by synthesizing the decoded differential image frame generated by the inverse orthogonal transformer 204 and the predicted image frame generated by the intra predictor 208 or the MC unit 206 selected by the switch 207. A predetermined deblocking process is performed on the decoded image frame obtained from 205 (step S39), thereby generating a decoded image frame after the deblocking process.

  Here, the predetermined deblocking process is performed according to the encoding mode of the frame or block currently being decoded, the state of the quantization parameter, the state of the image information in the surrounding decoded blocks, and the like. It is preferable that the block processing can be turned on / off and the strength of the filter processing can be switched. In addition, when the transform processing is correctly performed in the multi-harmonic local orthogonal transformer 121, it is generally determined that the deblocking process is not necessary, and the function is simply performed as an all-band pass filter without performing the deblocking process. It should be configured. The deblocking filter 209 supplies the generated decoded image frame after the deblocking process to the frame memory 210 so that the decoded image frame can be used as a reference image frame.

  The frame memory 210 acquires and stores the deblocked image frame from the deblocking filter 209 as a reference image frame. Further, the frame memory 210 reads the reference image frames accumulated as necessary, and supplies them to the MC unit 206 and the intra predictor 208. Here, it is desirable that the frame memory 210 can store at least one reference image frame and supply a necessary reference image frame according to the request of each component. Further, the frame memory 210 outputs a decoded image frame according to the display timing of the decoded image frame (step S40). An image frame necessary for display is output as an output code 211 to an external output display.

  Thereafter, it is further determined whether there is an encoded bit stream to be decoded (step S41). If decoding is necessary, the process returns to step S31 to continue the decoding process after step S31 again. If there is no encoded bitstream to be decoded, it is assumed that there is no need for decoding. A series of decryption processing ends.

  By performing encoding / decoding processing on a moving image frame based on the configuration as described above, in this embodiment, a boundary is generated when orthogonal transformation is performed on a block area in a moving image frame to be encoded. An approximate waveform that cancels out the discontinuity of the portion and improves the smoothness of the waveform at the boundary portion can be specified from the boundary condition between the blocks. Thereafter, a residual orthogonal transform coefficient can be generated from the difference from the original orthogonal transform coefficient. In the present embodiment, by generating such a residual orthogonal transform coefficient, it is possible to suppress generation of unnecessary high-frequency components that has been a problem in the conventional method.

  In addition, the generated residual orthogonal transform coefficient is subjected to predetermined quantization and predetermined entropy coding as an encoding target instead of the conventional orthogonal transform coefficient, thereby improving the coding efficiency in the conventional moving picture coding. it can.

  Next, another embodiment for realizing the encoding apparatus of the present invention will be described below with reference to FIGS. FIG. 5 is a block diagram showing another embodiment of the moving picture coding apparatus according to the present invention. In the figure, the same components as those in FIG. By the way, generally, when the prediction accuracy of motion estimation / motion compensation is not sufficiently high, a predicted image frame in which continuity is not always maintained between adjacent blocks in the frame is generated. A difference image frame including a residual component is generated from the difference between the predicted image frame and the original image frame to be encoded. Even in this difference image frame, continuity is always maintained between adjacent blocks. However, there is a case where the block boundary becomes discontinuous.

  When the conventional multiple harmonic local orthogonal transformation method is applied to such a difference image frame, the correlation between the blocks is not necessary even though it is necessary to perform the encoding so as to maintain the discontinuous state between the blocks. Encoding processing is performed on the assumption that the continuity is high and sufficient encoding efficiency may not be obtained as compared with a natural image. Also, if the amount of code required to reproduce the discontinuity between blocks at the time of encoding has been reduced by processing such as quantization, the decoding side performs decoding assuming continuity between blocks Therefore, discontinuity between blocks cannot be reproduced, and a problem arises in that this influence is propagated as an error by motion compensated prediction of moving picture decoding.

  In order to solve such a problem, the embodiment shown in FIG. 5 is characterized in that at least a boundary condition analyzer 131 and a boundary condition determiner 132 are further added to the encoding apparatus shown in FIG. is there. In FIG. 5, the orthogonal transformer 103 has a function of supplying the coefficient information after the orthogonal transformation to the boundary condition analyzer 131 instead of supplying the coefficient information to the multiple harmonic local orthogonal transformer 121. Here, the coefficient information after orthogonal transformation is DCT coefficient information. Since the other functions are the same as those in the configuration of FIG.

  The boundary condition analyzer 131 has a function of acquiring coefficient information after orthogonal transformation, here DCT coefficient information, from the orthogonal transformer 103. Further, the boundary condition analyzer 131 is used for a case where continuity is not always maintained between blocks, such as a predicted image frame of a moving image generated based on the MC processing or intra prediction processing used in the present invention. Analyzing the discontinuity of the image signal between the block to be encoded and the adjacent surrounding block, and determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition. In addition to having a function of determining, it has a function of generating boundary condition correction information for specifying whether correction of boundary conditions is necessary. The boundary condition correction information generated by the boundary condition analyzer 131 is output to the multiple harmonic local orthogonal transformer 121 and the boundary condition determiner 132, respectively.

  Further, the boundary condition analyzer 131 may have a function of supplying the generated boundary condition correction information to the entropy encoder 123 as necessary. By supplying the boundary condition correction information to the entropy encoder 123 in this way, even when it is difficult to estimate the block boundary state from the decoding result, the block boundary discontinuity can be uniquely specified on the decoding side. Such information can be obtained from the encoded bitstream.

  With respect to the determination condition when the boundary condition analyzer 131 determines whether or not the boundary condition needs to be corrected, it is generally noted that the inclination of the image signal at the block boundary, which is the block boundary condition, When the inclination of the image signal between the blocks is compared with the inclination of the image signal between the surrounding blocks, and the change in the inclination of the image signal exceeds a predetermined threshold, the inclination of the image signal is zero. If the change in the inclination of the image signal is within a predetermined threshold value, the inclination of the image signal between the blocks to be encoded is used as it is, and the multiharmonic local orthogonal transformer 121 in the subsequent stage is used. It is advisable to perform processing.

  In order to explain more specifically, the boundary condition analysis process and the multi-harmonic local orthogonal transform process of the block whose value is the encoding target are performed on the processing target as shown in FIGS. think of. 7 to 11, a moving image frame to be encoded is divided into predetermined block sizes, for example, each block size of 4 pixels in the horizontal direction and 4 pixels in the vertical direction (4 × 4) as an example. It is assumed that DCT coefficient information is obtained for each block by performing orthogonal transform processing, here, DCT processing.

  FIG. 7 is a conceptual diagram illustrating the positional relationship and block arrangement of each DCT coefficient information generated by DCT for a 4 × 4 block as an example. Here, it is assumed that the current block to be processed is the central block (corresponding to the second from the left and second from the top) 600 in FIG. Each block includes 4 × 4 DCT coefficient information. The white circle of each block is DCT coefficient information corresponding to the DC component of the entire block.

  The black circle in the horizontal direction is DCT coefficient information corresponding to a component obtained by DCTing the DC component in the vertical direction of the block in the horizontal direction. The black circles in the vertical direction are DCT coefficient information corresponding to components obtained by DCTing the DC component in the horizontal direction of the block in the vertical direction. The shaded portion of each block is DCT coefficient information corresponding to the remaining high frequency components.

  FIG. 8 is a conceptual diagram illustrating a process of performing multiple harmonic local orthogonal transform processing in the horizontal direction on the DCT coefficient information of the high-frequency component that is the shaded portion of the block in the central processing target block 600 in FIG. is there. Here, a process of performing the multiharmonic local orthogonal transform process on the DCT coefficient information corresponding to the first row of the hatched portion to be processed is shown. The DCT coefficient information of the inclination of the image signal whose boundary condition is the above-described equation (9) can be obtained approximately from the DCT coefficient information. Therefore, the DCT coefficient information of the inclination of the image signal which is the left boundary condition is obtained using the DCT coefficient information of a and b in FIG. Similarly, the right boundary condition is obtained using the DCT coefficient information of a and c.

  Thereafter, approximate DCT coefficient information, which is an estimated signal of DCT coefficient information corresponding to the first line of the hatched portion, is obtained using equation (1). Similarly, for the second and subsequent lines, the approximate DCT coefficient information for the second and subsequent lines is obtained using the DCT coefficient information at positions a and b and a and c corresponding to the second and subsequent lines. .

  FIG. 9 is a conceptual diagram for illustrating a process in which the same processing as in FIG. 8 is performed in the vertical direction. Since the processing itself is performed in the vertical direction in the same manner as the processing performed in the horizontal direction, the approximate DCT coefficient information in the vertical direction can be obtained, and thus description thereof is omitted here.

  FIG. 10 is a conceptual diagram when DCT coefficient information corresponding to a component obtained by DCT of a horizontal DC component of a block in the vertical direction is a processing target of multiple harmonic local orthogonal transformation. In the processing for the processing target in FIG. 10, the boundary condition may be approximately obtained by using DCT coefficient information that is the DC component of each entire block. Here, the DCT coefficient information of the inclination of the image signal, which is the upper boundary condition, is obtained using the DCT coefficient information of A and B. Similarly, the lower boundary condition is obtained using the DCT coefficient information of A and C. Thereafter, using the equation (1), the approximate DCT coefficient information of the region to be processed is obtained.

  FIG. 11 is a conceptual diagram for illustrating a process of performing the same processing as FIG. 10 in the horizontal direction. Since the processing performed in the vertical direction of FIG. 10 is similarly performed in the horizontal direction, approximate DCT coefficient information can be obtained, and thus description thereof is omitted here.

  FIG. 12 is a conceptual diagram for representing the positional relationship and block arrangement of each DCT coefficient information generated by 4 × 4 DCT as an example as in FIG. 7. In FIG. 12, the white circle of each block is DCT coefficient information corresponding to the DC component of the entire block. The black circle in the horizontal direction is DCT coefficient information corresponding to a component obtained by DCTing the DC component in the vertical direction of the block in the horizontal direction. The black circles in the vertical direction are DCT coefficient information corresponding to components obtained by DCTing the DC component in the horizontal direction of the block in the vertical direction. The shaded portion of each block is DCT coefficient information corresponding to the remaining high frequency components.

  Here, it is assumed that the current block to be processed is the central block (corresponding to the third from the left and third from the top) 610 in FIG. In FIG. 13 to FIG. 16, a moving image frame to be encoded is divided into predetermined block sizes, in this example, 4 × 4 block sizes, and orthogonal transform processing, here DCT processing, is performed on each block. It is assumed that DCT coefficient information is obtained for each block.

Under the above conditions, the analysis processing of the boundary condition of Γ ⁽³⁾ which is the boundary between blocks shown in FIG. 34 will be considered. Is usually only focus on adjacent blocks as shown in FIG. 8 is, as shown in FIG. 13, a reference block showing the block 610 which is a coded FIG 34 Omega ^(0,0) We are paying attention to the left two blocks. The coefficient information A in FIG. 13 is the DC component F _0,0 ^(0,0) of the entire block Ω ^(0,0) to be encoded. The coefficient information B is the DC component F _0,0 ^(-1 , ^{0) of the} entire block Ω ^(−1,0) one left from the encoding target block, and the coefficient information D is two left from the encoding target block. The DC component F _0,0 ^(−2, ₀ ^{) of the} entire block Ω ^(−2, ₀ ⁾ .

As shown in the equation ⁽⁹⁾ , the gradient G _k ⁽³⁾ , which is a boundary condition, can be approximately obtained from the DCT coefficient information F _{k1, k2,} so that G ₀ ⁽³ in Ω ^{(0,0) )} And G ₀ ^{(3 )} in Ω ^{(−1, 0)} are obtained from coefficient information A, coefficient information B, and coefficient information D. Thereafter, G ₀ ^{(3 )} in Ω ^(0,0) is compared with G ₀ ⁽³⁾ in Ω ^(−1,0) , and when the change in the inclination of the image signal exceeds a predetermined threshold value, Information that can determine that the boundary of Γ ⁽³⁾ is discontinuous, specify the discontinuity of the boundary portion by acquiring boundary condition correction information, and specify that the inclination of the image signal should be zero Generate.

If the change in the inclination of the image signal is within a predetermined threshold, the subsequent processing of the multiharmonic local orthogonal transformer 121 is performed using the inclination of the image signal between the blocks to be encoded as they are. It is good to do so. The boundary condition correction information is preferably configured to generate flag information such as continuous / discontinuous 0/1 for each boundary. By performing the above analysis process on the right boundary Γ ⁽⁴⁾ , the upper boundary Γ ⁽¹⁾ and the lower boundary Γ ⁽²⁾ in the same manner, the block Ω ^(0,0) The analysis process of the boundary condition can be advanced.

Here, in order to simplify the analysis processing, the boundary discontinuity is determined using only G _0, but if necessary, F _{0, s} like coefficient information a to e in FIG. ^{(T, 0)} {s = 1, 2,..., N-1, t = −2, −1, 0, 1, 2} G _k {k = 1, 2,. N-1} may be used to determine the discontinuity of the block Ω ^(0,0) boundary. This can be similarly applied to FIGS.

  Further, the boundary condition analyzer 131 shown in FIG. 5 performs the boundary condition analysis processing so that the boundary condition can be more accurately specified and the boundary discontinuity can be determined at the time of decoding. The quantizer 104 obtains a quantization parameter that fluctuates according to fixed or predetermined code amount control used when the quantization process is performed, and a DCT coefficient after quantization / inverse quantization based on the quantization parameter After estimating the information, the discontinuity of the image signal between the block to be encoded and the neighboring blocks adjacent thereto is analyzed, and the boundary condition is corrected based on at least one predetermined determination condition. It is better to have a function of determining whether it is necessary.

  The multi-harmonic local orthogonal transformer 121 in the configuration of FIG. 5 has a function of acquiring coefficient information (DCT coefficient information here) after orthogonal transformation via the orthogonal transformer 103 or the boundary condition analyzer 131, and boundary conditions. The function of acquiring boundary condition correction information from the analyzer 131, and controlling the inclination of the image signal that is the boundary condition of each block based on the acquired boundary condition correction information, and the Poisson equation based on the boundary condition of each block The residual DCT coefficient information is obtained from the difference between the function for obtaining approximate orthogonal transform coefficient information (here, approximate DCT coefficient information) in the block by analytically solving, and the obtained DCT coefficient information and the obtained approximate DCT coefficient information. And a function of notifying the quantizer 104 of the generated residual DCT coefficient information.

  Here, the inclination control of the image signal based on the acquired boundary condition correction information specifies the discontinuity of the boundary portion from the acquired boundary condition correction information, and the image signal when the target boundary portion is discontinuous. Correct the tilt. In general, it is desirable to correct the inclination of the image signal so that the inclination is zero. Further, when all the boundary portions are discontinuous, it is better to have a function of notifying the quantizer 104 of the acquired DCT coefficient information as it is without performing the multiple harmonic local orthogonal transform process.

  In addition, the multiple harmonic local orthogonal transformer 121 in the configuration of FIG. 5 performs orthogonal transformation (here, DCT is performed) on an estimation signal in a predetermined rectangular region that is an encoding target that satisfies the Poisson equation. When generating the obtained approximate DCT coefficient information using the DCT coefficient information after orthogonal transformation, a quantization parameter used for quantization is obtained for the DCT coefficient information after orthogonal transformation, and the quantization parameter is used as the quantization parameter. Based on the estimation result, the DCT coefficient information after quantization is estimated, and approximate DCT coefficient information is generated based on the estimation result.

  The quantizer 104 in the configuration of FIG. 5 has a better configuration if it has a function of notifying the boundary condition analyzer 131 and the multiharmonic local orthogonal transformer 121 of the quantization parameter used at the time of quantization. Since the other functions are the same as those in the configuration of FIG. 5 has a function of notifying the boundary condition determination unit 132 of the generated post-quantization information. Since the other functions are the same as those in the configuration of FIG.

  The boundary condition determining unit 132 has a function of acquiring boundary condition correction information from the boundary condition analyzer 131, a function of acquiring post-inverse quantization information from the inverse quantizer 105, and a case where boundary condition correction information can be acquired (valid If the boundary condition correction information cannot be acquired (if it is not valid) and the function for notifying the demultiplexed harmonic local orthogonal transformer 122 as the boundary condition determination information and the boundary condition correction information is acquired. Analyzes the discontinuity of the image signal between the block to be decoded and the neighboring blocks using the information after inverse quantization, and determines the boundary condition based on at least one predetermined determination condition. And a function of determining whether correction is necessary.

The boundary condition determiner 132 has a function of generating boundary condition correction information for specifying whether the boundary condition needs to be corrected and notifying the demultiplexed harmonic local orthogonal transformer 122 as boundary condition determination information. . Here, the boundary condition determination information generates information corresponding to G ₀ using the obtained inverse quantization information, that is, the DC component in the residual DCT coefficient information after inverse quantization, and the discontinuity between blocks. It is desirable to generate boundary condition determination information by performing the above determination. It would be even better if it had a function to generate boundary condition determination information based on such discontinuity determination. Further, the boundary condition determination unit 132 has a function of notifying the acquired dequantized information to the inverse multiple harmonic local orthogonal transformer 122.

  The inverse multiple harmonic local orthogonal transformer 122 in the configuration of FIG. 5 has a function of acquiring boundary condition determination information and post-inverse quantization information from the boundary condition determination unit 132. Here, the inclination control of the image signal based on the acquired boundary condition determination information specifies the discontinuity of the boundary part from the acquired boundary condition determination information, and the image signal when the target boundary part is discontinuous Correct the tilt. In general, it is desirable to correct the inclination of the image signal so that the inclination is zero. Since the other functions are the same as those in the configuration of FIG.

  The entropy encoder 123 in the configuration of FIG. 5 may have a function of acquiring boundary condition correction information notified from the boundary condition analyzer 131 as necessary. In this way, the entropy encoder 123 acquires the boundary condition correction information, so that even when it is difficult to estimate the block boundary state from the decoding result, the block boundary discontinuity can be uniquely specified on the decoding side. Such information can be incorporated into the encoded bitstream based on a predetermined syntax structure. Since the other functions are the same as those in the configuration of FIG.

  Next, the operation of another embodiment of the encoding apparatus of the present invention shown in FIG. 5 will be described with reference to the flowchart of FIG. First, an image frame to be encoded is prepared as the input code 101 of the encoding apparatus in FIG. 5 (step S71), and then it is determined whether or not intra encoding is performed (step S72). In the case of intra coding, intra prediction processing is performed by the intra predictor 110 (step S73). A predicted image frame obtained by intra prediction is output to the terminal a side of the switch 113 in FIG. Further, the switch 113 is switched to the terminal a side. Thereafter, the process proceeds to step S76.

  In the case of non-intra coding, the ME unit 111 in FIG. 5 acquires a moving image frame that is currently a coding target and a predetermined reference image frame stored in the frame memory 109, and performs motion estimation. (Step S74). The motion vector information generated by this motion estimation processing is supplied to the MC unit 112 and is also supplied to the entropy encoder 123 to be entropy encoded.

  The MC unit 112 acquires the motion vector information generated by the ME unit 111 through the motion estimation process, acquires the reference image frame used when the ME unit 111 generates the motion vector information from the frame memory 109, and moves these motions. Motion compensation is performed from the vector information and the reference image frame to generate a predicted image frame (step S75). This predicted image frame is output to the terminal b side of the switch 113. Further, the switch 113 is switched to the terminal b side, and then the process proceeds to step S76.

  The orthogonal transformer 103 in FIG. 5 performs prediction obtained from the moving image frame obtained as the input code 101 and the intra predictor 110 connected to the terminal a of the switch 113 or the MC unit 112 connected to the terminal b of the switch 113. A difference image frame with respect to the image frame is acquired from the subtractor 102, and predetermined orthogonal transformation is performed on the difference image frame (step S76). The orthogonal transformer 103 supplies the coefficient information after the orthogonal transformation to the boundary condition analyzer 131. Here, the coefficient information after orthogonal transform is DCT coefficient information.

  The boundary condition analyzer 131 acquires coefficient information after orthogonal transformation, here DCT coefficient information, from the orthogonal transformer 103, analyzes boundary conditions based on the DCT coefficient information (step S77), and boundary condition correction information. Is generated. Here, when the boundary condition is analyzed, the quantization parameter used by the quantizer 104 is acquired, and the coefficient information (here, DCT coefficient information) after orthogonal transformation is quantized based on the quantization parameter. DCT coefficient information after performing quantization / inverse quantization may be estimated, and boundary conditions may be analyzed based on the estimated DCT coefficient information.

  The generated boundary condition correction information is supplied to the multiple harmonic local orthogonal transformer 121 and the boundary condition determiner 132. Further, the boundary condition correction information may be supplied to the entropy encoder 123 as necessary. Further, the boundary condition analyzer 131 has a function of notifying the multiharmonic local orthogonal transformer 121 of the obtained coefficient information after orthogonal transformation (here, DCT coefficient information).

  The multiharmonic local orthogonal transformer 121 obtains coefficient information (DCT coefficient information in this case) and boundary condition correction information after orthogonal transformation from the boundary condition analyzer 131, and a block of interest based on the boundary condition correction information. The boundary part is analyzed to determine the discontinuity of the boundary condition. When the boundary part is discontinuous, the boundary condition is corrected, and then the multiple harmonic local orthogonal transformation process is performed (step S78). Here, when performing the multi-harmonic local orthogonal transform processing, the quantization parameter used by the quantizer 104 is acquired, and the coefficient information (here, DCT coefficient information) after the orthogonal transform is obtained based on the quantization parameter. Thus, the DCT coefficient information after the quantization / inverse quantization may be estimated, and the process may be performed based on the estimated DCT coefficient information. The generated residual DCT coefficient information is supplied to the quantizer 104.

  The quantizer 104 obtains residual DCT coefficient information from the multiple harmonic local orthogonal transformer 121, performs predetermined quantization on the residual DCT coefficient information, and uses the residual DCT coefficient information after quantization. Certain post-quantization information is generated (step S79). The generated post-quantization information is supplied to the inverse quantizer 105 and the entropy encoder 123.

  The inverse quantizer 105 obtains post-quantization information from the quantizer 104 and performs predetermined inverse quantization on the post-quantization information, thereby obtaining residual DCT coefficient information after inverse quantization. Information after inverse quantization is generated (step S80). The generated post-quantization information is supplied to the boundary condition determination unit 132.

  The boundary condition determiner 132 acquires post-inverse quantization information from the inverse quantizer 105, acquires boundary condition correction information from the boundary condition analyzer 131, and determines boundary conditions based on the acquired boundary condition correction information. Is performed (step S81). Here, when boundary condition correction information can be acquired, the acquired boundary condition correction information is supplied to the demultiplex harmonic local orthogonal transformer 122 as boundary condition determination information, and when boundary condition correction information cannot be acquired, acquisition is performed. Analyzing the discontinuity of the image signal between the block to be decoded and the adjacent surrounding block using the dequantized information, the boundary condition based on at least one predetermined determination condition It is determined whether or not correction is necessary. The boundary condition determiner 132 generates boundary condition correction information for specifying whether the boundary condition needs to be corrected, supplies the boundary condition determination information to the demultiplex harmonic local orthogonal transformer 122 as boundary condition determination information, and acquires the boundary condition correction information. The dequantized information is supplied to the inverse multiple harmonic local orthogonal transformer 122.

  The demultiplex harmonic local orthogonal transformer 122 acquires boundary condition determination information and dequantized information from the boundary condition determiner 132, and based on the boundary condition determination information and dequantized information, a predetermined demultiplex harmonic is obtained. A local orthogonal transform process is performed (step S82). Thereby, the demultiplex harmonic local orthogonal transformer 122 generates decoded DCT coefficient information obtained by decoding from the residual DCT coefficient information after dequantization, and the decoded DCT coefficient information is sent to the inverse orthogonal transformer 106. Supply.

  The inverse orthogonal transformer 106 acquires the decoded DCT coefficient information from the inverse multiple harmonic local orthogonal transformer 122, and performs inverse orthogonal transformation on the decoded DCT coefficient information (step S83). Thereby, the inverse orthogonal transformer 106 generates a decoded difference image frame, and supplies the decoded difference image frame to the intra predictor 110 and the deblocking filter 108 via the adder 107. The deblocking filter 108 adds a decoded image frame obtained by synthesizing the decoded difference image frame generated by the inverse orthogonal transformer 106 and the predicted image frame generated by the intra predictor 110 or the MC unit 112 selected by the switch 113 to the adder 107. A predetermined deblocking process is performed on the decoded image frame acquired as necessary (step S84).

  However, it is often determined that there is no need for the normal deblocking process, and it often functions simply as an all-band pass filter without performing the deblocking process. The deblocking filter 108 stores the generated decoded image frame after the deblocking process in the frame memory 109 so that it can be used as a reference image frame.

  The entropy encoder 123 acquires at least post-quantization information from the quantizer 104 and motion vector information from the ME unit 111, and acquires boundary condition correction information notified from the boundary condition analyzer 131 as necessary. You may make it do. The entropy encoder 123 performs predetermined entropy encoding on at least the obtained post-quantization information and motion vector information (step S85). The generated entropy-encoded information is supplied to the multiplexer 124.

  The multiplexer 124 obtains the entropy-encoded information from the entropy encoder 123 and multiplexes the encoded information based on a predetermined syntax structure to generate an encoded bitstream (step S86). . The generated encoded bit stream is output to the output transmitter 125. The output transmitter 125 acquires the encoded bit stream from the multiplexer 124, generates packet information based on a predetermined packetization process, and outputs the packet information as an output code 126 to a predetermined transmission path or a predetermined recording medium. .

  By sequentially performing the encoding process as described above on a moving image frame to be encoded, the encoding apparatus of the present invention uses an encoded bit stream having higher encoding efficiency than the conventional encoding apparatus as an output code. Obtainable.

  Next, another embodiment of the decoding apparatus of the present invention will be described below with reference to FIGS. FIG. 17 shows a block diagram of another embodiment of a decoding apparatus according to the present invention. In the figure, the same components as those in FIG. 17, the entropy decoder 224 obtains, from the demultiplexer 223, boundary condition correction information that is encoded and incorporated as necessary by the entropy encoder 123 of the encoding device of FIG. It is preferable to have a function of extracting from the information after the separation, and a function of notifying the acquired boundary condition correction information to the boundary condition determination unit 231. Since the other functions are the same as those in the configuration of FIG. The inverse quantizer 225 has a function of notifying the boundary condition determination unit 231 of the generated post-quantization information. Since the other functions are the same as those in the configuration of FIG.

  The boundary condition determination unit 231 has a function of acquiring boundary condition correction information from the entropy decoder 224, a function of acquiring post-quantization information from the inverse quantizer 225, and a case where boundary condition correction information can be acquired. A function for supplying the acquired boundary condition correction information to the demultiplex harmonic local orthogonal transformer 226 as boundary condition determination information, and if the boundary condition correction information cannot be acquired, decoding is performed using the acquired post-quantization information. A function that analyzes the discontinuity of an image signal between a block to be converted to an adjacent block and determines whether a boundary condition needs to be corrected based on at least one predetermined determination condition And have. Further, the boundary condition determination unit 231 has a function of generating boundary condition correction information for specifying whether the boundary condition needs to be corrected, and supplying the boundary condition correction information to the inverse multiple harmonic local orthogonal transformer 226 as boundary condition determination information. .

Here, the boundary condition determination information generates information corresponding to G ₀ using the obtained inverse quantization information, that is, the DC component in the residual DCT coefficient information after the inverse quantization, and the discontinuity between the blocks. It is desirable to perform boundary determination and generate boundary condition determination information. The boundary condition determiner 231 has a better configuration if it has a function of generating boundary condition determination information based on such a determination of discontinuity, and the obtained post-quantization information is demultiplexed with harmonic local orthogonality. A function of notifying the converter 226 is provided.

  The inverse multiple harmonic local orthogonal transformer 226 has a function of acquiring boundary condition determination information from the boundary condition determination unit 231. Since the other functions are the same as those in the configuration of FIG.

  Next, the operation of another embodiment of the decoding apparatus of the present invention shown in FIG. 17 will be described with reference to the flowchart of FIG. First, the input receiver 222 sequentially acquires predetermined packet information from a predetermined transmission path or a predetermined recording medium as an input code, reconstructs the encoded bit stream, and then multiplexes and separates the encoded bit stream To the container 223. The demultiplexer 223 acquires the encoded bit stream reconstructed by the input receiver 222, and performs demultiplexing of encoded information on the acquired encoded bit stream based on a predetermined syntax structure ( Step S71), the information after demultiplexing is supplied to the entropy decoder 224.

  The entropy decoder 224 obtains information after demultiplexing and performs predetermined entropy decoding, so that it is necessary to construct post-quantization information and motion vector information, and other predetermined syntax structures. Parameter information and the like are generated (step S72). The entropy decoder 224 may generate boundary condition correction information from the information after demultiplexing as necessary. Thereafter, at least post-quantization information is supplied to the inverse quantizer 225 and motion vector information is supplied to the MC unit 206. In addition, boundary information correction information may be supplied to the boundary information determination unit 231 as necessary.

  The inverse quantizer 225 performs predetermined inverse quantization on the post-quantization information acquired from the entropy decoder 224, thereby obtaining dequantized information that is residual DCT coefficient information after inverse quantization. Generate (step S73). The generated dequantized information is supplied to the boundary condition determiner 231. Thereafter, it is determined whether or not intra decoding is performed (step S74).

  In the case of intra decoding, intra prediction processing is performed by the intra predictor 208 (step S75), and a predicted image frame obtained by intra prediction is supplied to the terminal a side of the switch 207. Further, the switch 207 is switched to the terminal a side. Thereafter, the process proceeds to step S77.

  If it is determined in step S75 that it is not intra coding, the MC unit 206 acquires motion vector information from the entropy decoder 224 and also uses the reference image frame used when generating the acquired motion vector information. A reference image frame decoded and stored so as to constitute a reference image frame equivalent to the above is acquired from the frame memory 210, and motion compensation is performed from the acquired motion vector information and reference image frame (step S76). As a result, the MC unit 206 generates a predicted image frame and supplies it to the terminal b side of the switch 207. Further, the switch 207 is switched to the terminal b side. Thereafter, the process proceeds to step S77.

  Next, the boundary condition determination unit 231 acquires boundary condition correction information from the entropy decoder 224, acquires post-inverse quantization information from the inverse quantizer 225, and acquires the acquired boundary condition correction information and inverse quantum. Based on the converted information, the boundary condition is determined (step S77). When the boundary condition correction information can be acquired, the boundary condition determination unit 231 supplies the acquired boundary condition correction information to the inverse multiple harmonic local orthogonal transformer 226 as boundary condition determination information and cannot acquire the boundary condition correction information. Includes analyzing the discontinuity of an image signal between a block to be decoded and a neighboring block adjacent to the block after decoding using the obtained post-quantization information, and satisfying at least one predetermined determination condition Based on this, it is determined whether the boundary condition needs to be corrected. The boundary condition determiner 231 generates boundary condition correction information for specifying whether the boundary condition needs to be corrected, supplies the boundary condition determination information to the demultiplex harmonic local orthogonal transformer 226 as boundary condition determination information, and acquires the boundary condition correction information. The inversely quantized information is supplied to the inverse multiple harmonic local orthogonal transformer 226.

  The inverse multiple harmonic local orthogonal transformer 226 performs a predetermined inverse multiple harmonic local orthogonal transform based on the boundary condition determination information and the dequantized information acquired from the boundary condition determiner 231, thereby performing the post-inverse quantization. Decoded DCT coefficient information obtained by decoding from residual DCT coefficient information is generated (step S78). The generated decoded DCT coefficient information is supplied to the inverse orthogonal transformer 204.

  The inverse orthogonal transformer 204 acquires the decoded DCT coefficient information from the inverse multiple harmonic local orthogonal transformer 226, and generates a decoded difference image frame by performing inverse orthogonal transformation on the acquired decoded DCT coefficient information (step S79). ). The generated decoded difference image frame is supplied to the adder 205, where it is combined with the predicted image frame generated by the intra predictor 208 or the MC unit 206 selected by the switch 207 and generated as a decoded image frame. . The decoded image frame is supplied to the intra predictor 208 and the deblocking filter 209.

  The deblocking filter 209 acquires a decoded image frame obtained by combining the decoded difference image frame generated by the inverse orthogonal transformer 204 and the predicted image frame generated by the intra predictor 208 or the MC unit 206 selected by the switch 207. Then, by performing a predetermined deblocking process on the decoded image frame, a decoded image frame after the deblocking process is generated (step S80).

  Here, the predetermined deblocking process is performed according to the encoding mode of the frame or block currently being decoded, the state of the quantization parameter, the state of the image information in the surrounding decoded blocks, and the like. It is preferable that the block processing can be turned on / off and the strength of the filter processing can be switched. In addition, when the inverse multiharmonic local orthogonal transformer 226 performs the conversion process correctly, it is generally determined that the deblocking process is not necessary, and the demultiplexing process is not performed, and it simply functions as an all-band pass filter. It is good to be configured. The deblocking filter 209 supplies and stores the generated decoded image frame after the deblocking process to the frame memory 210 so that the decoded image frame can be used as a reference image frame.

  The frame memory 210 acquires and accumulates the deblocked image frame from the deblocking filter 209 as a reference image frame, and supplies the reference image frame to the MC unit 206 and the intra predictor 208 as necessary. Here, it is desirable that the frame memory 210 can store at least one reference image frame and supply a necessary reference image frame according to the request of each component. Also, the decoded image frame is output according to the display timing of the decoded image frame (step S81). The decoded image frame from the frame memory 210 is output to an external output display as an output code 211 that is an image frame necessary for display.

  Thereafter, it is further determined whether there is an encoded bit stream to be decoded (step S82). If there is an encoded bitstream to be decoded, the process returns to step S71, and the decoding process in steps S71 to S82 is continued again. If there is no encoded bitstream to be decoded, A series of decryption processing ends.

  In the case where continuity between blocks is not always maintained by performing encoding / decoding processing on a moving image frame based on the configuration as described above, the boundary between blocks to be encoded This method compares the conditions with the boundary conditions of the surrounding blocks, adopts the one with the smaller slope of the image signal that is the boundary condition of the block as the boundary condition of the block to be encoded, and the multiple harmonic local cosine transform method Apply. By performing such boundary condition determination of the encoding target block, the discontinuity between the blocks is correctly reproduced while improving the encoding efficiency, and this effect is caused as an error by the motion compensation prediction of the video decoding. The problem of propagation can be improved.

  The present invention relates to the central processing control device which is the computer shown in FIGS. 19, 20, 21, and 22 with the functions of the encoding device and the decoding device shown in FIGS. 1, 3, 5, and 17 described above. It includes a program for realizing 1903, 2003, 2103, 2203. This program may be taken in from a recording medium or may be taken in via a communication network.

  FIG. 19 is a block diagram showing an example of an information processing apparatus that operates according to an encoding program according to the present invention. In the figure, an information processing device 1900 operates according to an embodiment of an input device 1901 for inputting various types of information, an output device 1902 for outputting various types of information, and an encoding program of the present invention. A central processing control device 1903, an external storage device 1904, a temporary storage device 1905 used as a work area for arithmetic processing by the central processing control device 1903, and a communication device 1906 for communicating with the outside are provided with a bidirectional bus. 1907 is connected.

  The central processing control device 1903 receives the encoding program of the present invention from a recording medium or via a communication network and is taken in by the communication device 1906, and includes orthogonal transform means 1908 corresponding to the orthogonal transformer 103 in FIG. Multiple harmonic local orthogonal transform means 1909 corresponding to the harmonic local orthogonal transformer 121, quantization means 1910 corresponding to the quantizer 104, inverse quantization means 1911 corresponding to the inverse quantizer 105, inverse multiple harmonic local orthogonal transformer Inverse multiple harmonic local orthogonal transform unit 1912 corresponding to 122, inverse orthogonal transform unit 1913 corresponding to the inverse orthogonal transformer 106, intra prediction unit 1914 corresponding to the intra predictor 110, ME means 1915 corresponding to the ME unit 111, MC MC means 1916 corresponding to the device 112 and a de-blocking equivalent to the deblocking filter 108 A function corresponding to the output transmitter 125 by controlling the locking filter means 1917, the entropy encoding means 1918 corresponding to the entropy encoder 123, the multiplexing means 1919 corresponding to the multiplexer 124, the output device 1902 and the communication device 1906. And at least the functions of the output transmission means 1920 to realize the same operation as that of the encoding apparatus of the present invention shown in FIG. 1 by software processing.

  FIG. 20 is a block diagram showing an example of an information processing apparatus that operates according to the decryption program of the present invention. In the figure, an information processing device 2000 includes an input device 2001 for inputting various types of information, an output device 2002 for outputting various types of information, and a central processing control device 2003 that operates according to the decoding program of the present invention. And an external storage device 2004, a temporary storage device 2005 used for a work area in arithmetic processing by the central processing control device 2003, and a communication device 2006 for communicating with the outside are connected by a bidirectional bus 2007. It is configured.

  The central processing control device 2003 receives the decoding program of the present invention from a recording medium or via a communication network and is taken in by the communication device 2006, and entropy decoding means 2008 corresponding to the entropy decoder 224 of FIG. , Inverse quantization means 2009 corresponding to inverse quantizer 225, inverse multiple harmonic local orthogonal transform means 2010 corresponding to inverse multiple harmonic local orthogonal transformer 226, inverse orthogonal transform means 2011 equivalent to inverse orthogonal transformer 204, intra An intra prediction unit 2012 corresponding to the predictor 208, a deblocking filter unit 2013 corresponding to the deblocking filter 209, an MC unit 2014 corresponding to the MC unit 206, and an input receiver 222, the input device 2001 and the communication device 2006 Functions corresponding to the input receiver 222 by controlling It has at least the functions of the demultiplexing means 2016 corresponding to the input receiving means 2015 and the demultiplexing means 223, and performs the same operation as the decoding apparatus of the present invention shown in FIG. To do.

  FIG. 21 shows a configuration diagram of another example of the information processing apparatus that operates according to the encoding program of the present invention. In the figure, the same components as those in FIG. In FIG. 21, an information processing apparatus 2100 includes an input apparatus 2101 for inputting various types of information, an output apparatus 2102 for outputting various types of information, and a central processing control apparatus 2103 that operates according to the encoding program of the present invention. A two-way bus 2107 is connected to the external storage device 2104, a temporary storage device 2105 used for a work area in the arithmetic processing by the central processing control device 2103, and a communication device 2106 for communicating with the outside. It is configured.

  The central processing control device 2103 receives the encoding program of the present invention from a recording medium or via a communication network and is taken in by the communication device 2106, and includes orthogonal transform means 1908 corresponding to the orthogonal transformer 103 in FIG. Multiple harmonic local orthogonal transform means 1909 corresponding to the harmonic local orthogonal transformer 121, quantization means 1910 corresponding to the quantizer 104, inverse quantization means 1911 corresponding to the inverse quantizer 105, inverse multiple harmonic local orthogonal transformer Inverse multiple harmonic local orthogonal transform unit 1912 corresponding to 122, inverse orthogonal transform unit 1913 corresponding to the inverse orthogonal transformer 106, intra prediction unit 1914 corresponding to the intra predictor 110, ME means 1915 corresponding to the ME unit 111, MC MC means 1916 corresponding to the device 112 and a de-blocking equivalent to the deblocking filter 108 A function corresponding to the output transmitter 125 by controlling the locking filter means 1917, the entropy encoding means 1918 corresponding to the entropy encoder 123, the multiplexing means 1919 corresponding to the multiplexer 124, the output device 2102 and the communication device 2106. Output transmission means 1920, boundary condition analysis means 2108 corresponding to the boundary condition analyzer 131, and boundary condition determination means 2109 corresponding to the boundary condition determiner 132, at least. The same operation as that of the illustrated encoding apparatus of the present invention is executed by software processing.

  FIG. 22 shows a configuration diagram of another example of the information processing apparatus that operates according to the decryption program of the present invention. In the figure, the same components as in FIG. In FIG. 22, an information processing device 2200 includes an input device 2201 for inputting various types of information, an output device 2202 for outputting various types of information, and a central processing control device 2203 that operates according to the decoding program of the present invention. And an external storage device 2204, a temporary storage device 2205 that is used as a work area for arithmetic processing by the central processing control device 2203, and a communication device 2206 for communicating with the outside are connected by a bidirectional bus 2207. It is configured.

  The central processing control device 2203 is the entropy decoding means 2008 corresponding to the entropy decoder 224 of FIG. 17 by the decoding program of the present invention being distributed from a recording medium or distributed via a communication network and taken in by the communication device 2206. , Inverse quantization means 2009 corresponding to inverse quantizer 225, inverse multiple harmonic local orthogonal transform means 2010 corresponding to inverse multiple harmonic local orthogonal transformer 226, inverse orthogonal transform means 2011 equivalent to inverse orthogonal transformer 204, intra An intra prediction unit 2012 corresponding to the predictor 208, a deblocking filter unit 2013 corresponding to the deblocking filter 209, an MC unit 2014 corresponding to the MC unit 206, an input receiver 222, an input device 2201 and a communication device 2206 Functions controlled and equivalent to the input receiver 222 FIG. 17 shows at least the functions of the input receiving means 2015, the demultiplexing means 2016 corresponding to the demultiplexer 223, and the boundary condition judging means 2208 corresponding to the boundary condition judging unit 231 for realizing the above. The same operation as that of the decoding device of the present invention is executed by software processing.

  Next, a second embodiment of the present invention will be described. FIG. 23 shows a functional block diagram of the second embodiment of the moving picture coding apparatus according to the present invention. The present embodiment has a configuration in which the intra predictor 110 and the deblocking filter 108 are excluded from the moving picture coding apparatus of the embodiment of FIG. Each component of the video encoding device in FIG. 23 is the same as the corresponding component of the video encoding device according to the embodiment shown in FIG. 1, and a description thereof will be omitted here.

  This embodiment is characterized by intra coding, and if normal intra coding is performed on a picture including a signal component that is discontinuous and has a sudden change, sufficient coding efficiency cannot always be obtained, In addition, when selecting a block closest to the encoding target block from among the prediction blocks in intra prediction by AVC, it is not always possible to select a prediction block that is well approximated, so that sufficient encoding efficiency cannot be obtained. This is an embodiment for solving the above-described problem.

  Next, the operation of the moving picture coding apparatus of FIG. 23 will be described according to the flowchart of FIG. First, an image frame to be encoded is prepared as an input code 701 (step S71), and it is determined whether or not intra encoding is performed (step S72). In the case of intra coding, after the switch 715 is switched to the terminal a side, the process proceeds to step S75.

  On the other hand, if it is determined in step S72 that the encoding is not intra coding, the ME device 713 obtains a moving image frame that is the current encoding target and a predetermined reference image frame stored in the frame memory 712. Then, motion estimation is performed (step S73). Further, the MC unit 714 acquires the motion vector information generated by the ME unit 713 by the motion estimation process, acquires the reference image frame used when the ME unit 111 generates the motion vector information from the frame memory 712, and these (Step S74), and outputs to the terminal b side of the switch 715. Further, the switch 715 is switched to the terminal b side. Thereafter, the process proceeds to step S75.

  Since the operation in the case of non-intra coding is the same as that in the embodiment of FIG. 1, the description thereof is omitted here. Hereinafter, an operation in the case of intra coding will be described. The orthogonal transformer 703 acquires an image frame obtained as the input code 701, and performs predetermined orthogonal transformation on the acquired image frame (step S76). In the case of intra coding, since the switch 715 is connected to the terminal a, that is, disconnected so as not to obtain a prediction frame, an image frame obtained as an input code 701 instead of a difference frame is input to the orthogonal transformer 703. Note that The orthogonal transformer 703 notifies the boundary condition analyzer 704 of the coefficient information after the orthogonal transformation. Here, the coefficient information after orthogonal transform is DCT coefficient information.

  The boundary condition analyzer 704 obtains coefficient information (here, DCT coefficient information) after the orthogonal transformation from the orthogonal transformer 703, analyzes the boundary condition based on the obtained DCT coefficient information (step S76), Condition correction information is generated. Normally, boundary conditions between blocks adjacent to the current encoding target in the vertical and horizontal directions are analyzed, and necessary boundary condition correction information is generated for each boundary. Here, when the boundary condition is analyzed, the quantization parameter used by the quantizer 706 is acquired, and the coefficient information (here, DCT coefficient information) after orthogonal transformation is obtained based on the quantization parameter. The DCT coefficient information after the quantization / inverse quantization may be estimated, and the boundary condition may be analyzed based on the estimated DCT coefficient information.

  The boundary condition analyzer 704 notifies the generated boundary condition correction information to the multiple harmonic local orthogonal transformer 705 and the boundary condition determiner 708. Further, boundary condition correction information generated as necessary may be notified to the entropy encoder 716. Further, the boundary condition analyzer 704 notifies the multiharmonic local orthogonal transformer 705 of the obtained coefficient information after orthogonal transformation (here, DCT coefficient information).

  The multi-harmonic local orthogonal transformer 705 acquires coefficient information (here, DCT coefficient information) after orthogonal transformation and boundary condition correction information from the boundary condition analyzer 704, and a block of interest based on the acquired boundary condition correction information When the boundary portion is discontinuous, the boundary condition is corrected, and then the multiple harmonic local orthogonal transformation process is performed (step S77). Here, when performing the multiharmonic local orthogonal transform process, the quantization parameter used by the quantizer 706 is acquired, and the coefficient information (here, DCT coefficient information) after the orthogonal transform is obtained based on the quantization parameter. Alternatively, the DCT coefficient information after quantization / inverse quantization may be estimated, and processing may be performed based on the estimated DCT coefficient information. After that, the multiharmonic local orthogonal transformer 705 notifies the later-described quantizer 706 of the residual DCT coefficient information generated by the multiharmonic local orthogonal transform process.

  The quantizer 706 acquires residual DCT coefficient information from the multiharmonic local orthogonal transformer 705, performs predetermined quantization on the acquired residual DCT coefficient information (step S78), and performs the residual after quantization. Generate post-quantization information which is DCT coefficient information. The quantizer 706 notifies the generated post-quantization information to the inverse quantizer 706 and the entropy encoder 716.

  The inverse quantizer 706 obtains post-quantization information from the quantizer 706, performs predetermined inverse quantization on the obtained post-quantization information (step S79), and performs a residual DCT coefficient after inverse quantization. Information after dequantization that is information is generated. The inverse quantizer 706 notifies the generated post-quantization information to the boundary condition determiner 708.

  The boundary condition determiner 708 acquires post-inverse quantization information from the inverse quantizer 706, acquires boundary condition correction information from the boundary condition analyzer 704, and determines boundary conditions based on the acquired boundary condition correction information. This is performed (step S80). Here, when the boundary condition correction information can be acquired, the boundary condition determination unit 708 notifies the inverse multiple harmonic local orthogonal transformer 709 of the acquired boundary condition correction information as boundary condition determination information. If the boundary condition correction information cannot be acquired, the boundary condition determiner 708 uses the acquired post-quantization information to obtain the image signal between the block to be decoded and the adjacent surrounding blocks. The discontinuity is analyzed, and it is determined whether the boundary condition needs to be corrected based on at least one predetermined determination condition.

  Further, the boundary condition determiner 708 generates boundary condition correction information for specifying whether the boundary condition needs to be corrected, and notifies the demultiplex harmonic local orthogonal transformer 709 as boundary condition determination information. Usually, boundary conditions between blocks adjacent to the upper, lower, left, and right sides of the current encoding target are analyzed, and necessary boundary condition correction information is generated for each boundary. The boundary condition determiner 708 notifies the acquired dequantized information to the inverse multiple harmonic local orthogonal transformer 709.

  The inverse multiple harmonic local orthogonal transformer 709 acquires boundary condition determination information and dequantized information from the boundary condition determiner 708, respectively, and performs predetermined inverse based on the acquired boundary condition determination information and dequantized information. Multiple harmonic local orthogonal transform processing is performed (step S81). Thereby, the demultiplex harmonic local orthogonal transformer 709 generates decoded DCT coefficient information obtained by decoding from the residual DCT coefficient information after dequantization, and the generated decoded DCT coefficient information is converted into the inverse orthogonal transformer 710. Notify

  The inverse orthogonal transformer 710 obtains the decoded DCT coefficient information from the inverse multiple harmonic local orthogonal transformer 709, performs inverse orthogonal transformation on the obtained decoded DCT coefficient information (step S82), and generates a decoded image frame. The inverse orthogonal transformer 710 supplies the generated decoded image frame to the frame memory 712 via the adder 711 so that the decoded image frame can be used as a reference image frame.

  In the case of intra coding, the entropy encoder 716 obtains at least post-quantization information from the quantizer 706, and receives boundary condition correction information notified from the boundary condition analyzer 704 as necessary. You may make it acquire. The entropy encoder 716 performs predetermined entropy encoding on at least the acquired post-quantization information (step S83) and notifies the generated post-entropy encoding information to the multiplexer 717.

  The multiplexer 717 acquires the entropy-encoded information from the entropy encoder 716 and multiplexes the encoded information based on a predetermined syntax structure (step S84), thereby generating an encoded bitstream. . The generated encoded bit stream is notified to the output transmitter 718. The output transmitter 718 acquires the encoded bit stream from the multiplexer 717, generates packet information based on a predetermined packetization process, and outputs the packet information as an output code 719 to a predetermined transmission path or a predetermined recording medium. .

  By performing the encoding process as described above sequentially on a moving image frame to be encoded, the encoding apparatus according to the present embodiment performs encoding efficiency when performing intra-encoding compared to a conventional encoding apparatus. Can be obtained as an output code.

  In the above description of the moving picture encoding apparatus in FIG. 23, the encoding process is performed in units of frames. However, the present invention is not limited to this, and the present invention is not limited to this. You may comprise so that it may perform. Further, a configuration may be adopted in which a plurality of macroblocks are performed in units of slices. For example, in AVC (H.264), a macroblock (MB) can handle 16 horizontal pixels and 16 vertical pixels (hereinafter referred to as 16 × 16, etc.). Type), and can be divided into 16 × 16, 16 × 8, 8 × 16, and 8 × 8 areas according to the MB type. Each area is usually called a block. Further, when the MB mode is 8 × 8, there may be a plurality of types of blocks (Sub MB types), and 8 × 8, 8 × 4, 4 × 8, and 4 depending on the Sub MB type. It can be divided into x4 regions and handled. Each area is usually called a sub-block. By using such MB mode and Sub MB mode to perform motion estimation and motion compensation using various division methods, the conventional MPEG-1, -2, -4 (non-AVC) can be obtained. The estimation accuracy can be improved. Further, the unit of processing when performing integer DCT is 4 × 4.

  Therefore, next, the encoding process of the moving image encoding apparatus of FIG. 23 is performed in units of macroblocks, and the processing unit when performing the integer DCT that is orthogonal transform, multiple harmonic local orthogonal transform, or the like is 4 × 4. The operation will be described with reference to the flowchart of FIG. First, an image frame to be encoded is prepared as an input code 701 (step S71). Here, since the processing unit of integer DCT, which is orthogonal transform, is performed in units of 4 × 4 subblocks, first, the macroblock to be encoded is set as the upper left macroblock of the image frame. In this example, the encoding process is moved to the right macroblock from the upper left one by one in the raster order, and when the right end of the line is reached, it is sequentially processed while moving to the leftmost macroblock of the next line. And proceed with the story as if it were in the order of movement until it reached the bottom right macroblock. In addition, the processing order of the blocks in the macro block is in the order of upper left, upper right, lower left, and lower right, and the processing order of the sub-blocks in the block is assumed to be in the order of upper left, upper right, lower left, and lower right. To proceed. Such an encoding target macroblock is notified to the orthogonal transformer 703.

  Next, it is determined whether the macroblock that is the current encoding target is intra-encoded (step S72). In the case of intra coding, the switch 715 is switched to the terminal a side. Thereafter, the process proceeds to step S75. Since the operation in the case of non-intra coding is the same as that in the embodiment of FIG. 1, the description thereof is omitted here. Hereinafter, an operation in the case of intra coding will be described.

  The orthogonal transformer 703 acquires a macroblock that is a current encoding target of an image frame obtained as the input code 701. In the case of intra coding, the switch 715 is connected to the terminal a, that is, disconnected so as not to obtain a prediction block, and thus is a current coding target of an image frame obtained as an input code 701 instead of a difference block. Note that the macroblock becomes the input of the orthogonal transformer 703. Predetermined orthogonal transformation is performed for each 4 × 4 sub-block on the acquired macroblock (step S75). The orthogonal transformer 703 notifies the boundary condition analyzer 704 of the coefficient information after the orthogonal transformation. Here, the coefficient information after orthogonal transform is DCT coefficient information.

  The boundary condition analyzer 704 obtains coefficient information after orthogonal transformation in the macroblock to be encoded, that is, DCT coefficient information here, from the orthogonal transformer 703. Here, it is desirable to hold macroblock information that has undergone orthogonal transformation in advance. Furthermore, it is desirable to retain information on the orthogonally transformed sub-blocks adjacent to the upper and left sides of the sub-blocks to be encoded so that they can be used for analysis of boundary conditions. If further discontinuity analysis is required, two orthogonal transformed sub-blocks close to the top of the sub-block to be encoded and two orthogonal transformed adjacent to the left It is possible to obtain a better configuration by storing the sub-block information so that it can be used for boundary condition analysis. The boundary condition analyzer 704 analyzes the boundary condition based on the acquired DCT coefficient information (step S76), and generates boundary condition correction information. Here, when the boundary condition is analyzed, the quantization parameter used by the quantizer 706 is acquired, and the coefficient information (here, DCT coefficient information) after orthogonal transformation is quantized based on the quantization parameter. DCT coefficient information after performing quantization / inverse quantization may be estimated, and boundary conditions may be analyzed based on the estimated DCT coefficient information.

  Also, in the intra coding of the present invention, the coding order of each macroblock is coded in the horizontal direction from the macroblock in the first row on the upper left, and the coding proceeds in the second row and the third row sequentially. When encoding in a certain raster order, the boundary condition is analyzed only for the adjacent sub-blocks that have been encoded, and the boundary condition with the adjacent sub-block before encoding is analyzed without any analysis. It is advisable to perform the correction. In other words, at least analyze the boundary condition between the adjacent upper and left subblocks that have undergone orthogonal transformation, and do not analyze the boundary condition between the lower and right adjacent subblocks before orthogonal transformation. In addition, it is possible to obtain a better configuration by always correcting the boundary conditions. In the same macroblock, when the subblocks adjacent to the lower side and the right side of the subblock to be encoded have been orthogonally transformed, the boundary between the subblocks is the same as the upper side and the left side. By performing the condition analysis, the configuration is further improved.

  The boundary condition analyzer 704 notifies the generated boundary condition correction information to the multiple harmonic local orthogonal transformer 705 and the boundary condition determiner 708. In addition, boundary condition correction information may be notified to the entropy encoder 716 as necessary. Here, only the information obtained from the boundary condition with the sub-block after the orthogonal transformation is notified as the boundary condition correction information to be notified, and the boundary condition with the sub-block before the orthogonal transformation is always corrected. By doing so, the amount of information to be notified can be reduced, and a better configuration can be obtained. Further, the boundary condition analyzer 704 notifies the multiharmonic local orthogonal transformer 705 of the obtained coefficient information after orthogonal transformation (here, DCT coefficient information).

  The multiharmonic local orthogonal transformer 705 acquires coefficient information (here, DCT coefficient information) and boundary condition correction information after orthogonal transformation in the sub-block to be encoded from the boundary condition analyzer 704 and acquires the boundary condition correction information. Based on the boundary condition correction information, the discontinuity of the boundary portion of the sub-block of interest is determined. If the boundary portion is discontinuous, the boundary condition is corrected, and then the multiple harmonic local orthogonal transformation process is performed ( Step S77). Here, when performing the multiharmonic local orthogonal transform process, the quantization parameter used by the quantizer 706 is acquired, and the coefficient information (here, DCT coefficient information) after the orthogonal transform is obtained based on the quantization parameter. Alternatively, the DCT coefficient information after quantization / inverse quantization may be estimated, and processing may be performed based on the estimated DCT coefficient information.

  In addition, in the case of intra coding, the determination of the discontinuity of the boundary portion of the target sub-block, which is performed based on the notified boundary condition correction information, is performed only at the boundary portion with the sub-block that has undergone orthogonal transformation If the boundary portion with the sub-block before orthogonal transformation is determined to be discontinuous and the boundary condition is always corrected, the configuration is further improved. That is, the upper and left sub-blocks of the sub-block to be encoded are sub-blocks that have undergone orthogonal transformation, and thus the discontinuity of this boundary portion is determined. Since the lower and right sub-blocks of the sub-block to be encoded are sub-blocks before the orthogonal transformation, they are always discontinuous without determining the discontinuity of these boundary parts. It is desirable to perform processing as However, when the lower and right sub-blocks are in the same macroblock and are orthogonally transformed sub-blocks, it is desirable to determine the discontinuity of the boundary portion as in the upper and left sides. This is because the processing load is reduced by such processing, and higher-speed encoding and decoding can be realized. Thereafter, the multiple harmonic local orthogonal transformer 705 notifies the generated residual DCT coefficient information to the quantizer 706 described later.

  The quantizer 706 acquires residual DCT coefficient information from the multiharmonic local orthogonal transformer 705, performs predetermined quantization on the acquired residual DCT coefficient information (step S78), and performs the residual after quantization. Generate post-quantization information which is DCT coefficient information. The generated post-quantization information is notified to the inverse quantizer 706 and the entropy encoder 716.

  The inverse quantizer 706 obtains post-quantization information from the quantizer 706, performs predetermined inverse quantization on the obtained post-quantization information (step S79), and performs a residual DCT coefficient after inverse quantization. Information after dequantization that is information is generated. The generated post-quantization information is notified to the boundary condition determiner 708.

  The boundary condition determination unit 708 acquires post-inverse quantization information from the inverse quantizer 706. Also, boundary condition correction information is acquired from the boundary condition analyzer 704. A boundary condition is determined based on the acquired boundary condition correction information (step S80). Here, when the boundary condition correction information can be acquired, the boundary condition determination unit 708 notifies the inverse multiple harmonic local orthogonal transformer 709 of the acquired boundary condition correction information as boundary condition determination information. If the boundary condition correction information cannot be acquired, the boundary condition determination unit 708 uses the acquired post-quantization information to obtain an image between the sub-block to be decoded and the neighboring sub-blocks adjacent thereto. The signal discontinuity is analyzed to determine whether the boundary condition needs to be corrected based on at least one predetermined determination condition. Further, the boundary condition determiner 708 generates boundary condition correction information for specifying whether the boundary condition needs to be corrected, and notifies the demultiplex harmonic local orthogonal transformer 709 as boundary condition determination information.

  Here, when intra coding is performed in raster order, the boundary condition is analyzed only for the adjacent sub-blocks that have been encoded, and the boundary condition with the adjacent sub-block before encoding is analyzed. It is preferable to correct the boundary condition without performing it. In addition, the boundary condition determination unit 708 notifies the acquired dequantized information to the inverse multiple harmonic local orthogonal transformer 709.

  The inverse multiple harmonic local orthogonal transformer 709 acquires boundary condition determination information and dequantized information from the boundary condition determiner, and performs predetermined demultiplex harmonic based on the acquired boundary condition determination information and dequantized information. A local orthogonal transform process is performed (step S81), and decoded DCT coefficient information obtained by decoding from the residual DCT coefficient information after inverse quantization is generated. The generated decoded DCT coefficient information is notified to the inverse orthogonal transformer 710.

  The inverse orthogonal transformer 710 obtains the decoded DCT coefficient information from the inverse multiple harmonic local orthogonal transformer 709, performs inverse orthogonal transformation on the obtained decoded DCT coefficient information (step S82), and generates a decoded image block. The inverse orthogonal transformer 710 supplies the generated decoded image block to the frame memory 712 via the adder 711 and stores it in the corresponding position so that the generated decoded image block can be used as a reference image frame.

  In the case of intra coding, the entropy encoder 716 acquires at least post-quantization information from the quantizer 706. Further, boundary condition correction information notified from the boundary condition analyzer 704 may be acquired as necessary. The entropy encoder 716 performs predetermined entropy encoding on at least the acquired post-quantization information (step S83), and notifies the multiplexer 717 of the generated post-entropy encoding information.

  The multiplexer 717 acquires the entropy-encoded information from the entropy encoder 716, multiplexes the encoded information based on a predetermined syntax structure (step S84), and generates an encoded bitstream. The generated encoded bit stream is supplied to an output transmitter 718, where it is converted into packet information based on a predetermined packetization process, and then output as an output code 719 to a predetermined transmission path or a predetermined recording medium. The

  By performing the encoding process as described above sequentially on each block of a moving image frame to be encoded, the encoding apparatus of the present invention performs encoding when performing intra encoding more than a conventional encoding apparatus. An efficient encoded bit stream can be obtained as an output code. That is, in this embodiment, instead of performing intra prediction in normal intra coding, orthogonal transform is performed in units of a predetermined block on a frame that is a coding target of intra coding, and a coding target By analyzing the boundary conditions between this block and its neighbors and surrounding blocks, the correlation level between neighboring blocks is analyzed. If the correlation between neighboring blocks is high, , A predetermined multi-harmonic local orthogonal transformation is performed based on the boundary condition of adjacent blocks, a residual orthogonal transform coefficient is generated, and if the correlation between adjacent blocks is low as a result of the boundary condition analysis, By performing this conversion process after correcting the boundary conditions used in the orthogonal transform, it is possible to calculate the residual error generated by performing the process while considering discontinuities between blocks. The bias of the signal components of the transform coefficients, and larger than the intra-prediction used in conventional intra-coding, it is possible to obtain sufficient encoding efficiency.

  In the present embodiment, when encoding processing is performed in units of predetermined blocks, orthogonal transformation is performed when analyzing boundary conditions between a block to be encoded and its adjacent and surrounding blocks. Analyzing boundary conditions only with existing blocks and interpreting them as discontinuous with other blocks, reduces the processing of this embodiment and realizes faster encoding be able to.

  Furthermore, according to the present embodiment, the signal component in the encoding target block is subjected to multiple harmonic local orthogonal transformation, and the signal component is approximated by estimation based on the boundary condition with the adjacent block. Since the residual orthogonal transform coefficient is generated, the prediction block generated by the predetermined direction and the predetermined weight is selected as the closest to the encoding target block, as in the case of the conventional intra prediction. However, since the residual orthogonal transform coefficient can be generated based on a more complicated estimation signal, higher encoding efficiency can be obtained.

  Next, a second embodiment of the moving picture decoding apparatus of the present invention will be described. FIG. 25 shows a functional block diagram of the second embodiment of the moving picture decoding apparatus according to the present invention. Compared with the moving picture decoding apparatus shown in FIG. 3, the moving picture decoding apparatus of this embodiment has a configuration excluding the intra predictor 208 and the deblocking filter 209. Each component of the video decoding device in FIG. 25 is the same as the corresponding component of the video decoding device in the embodiment shown in FIG.

  Next, the operation of the video decoding apparatus in FIG. 25 will be described according to the flowchart in FIG. First, the input receiver 802 sequentially acquires predetermined packet information from a predetermined transmission path or a predetermined recording medium as an input code 801, and reconstructs an encoded bit stream. Thereafter, the demultiplexer 803 is notified of the encoded bit stream.

  The demultiplexer 803 acquires the encoded bitstream reconstructed by the input receiver 802, and demultiplexes the encoded information based on a predetermined syntax structure on the acquired encoded bitstream ( In step S91), the information after demultiplexing is notified to the entropy decoder 804.

  The entropy decoder 804 obtains information after demultiplexing and performs predetermined entropy decoding (step S92), thereby configuring post-quantization information, motion vector information, and other predetermined syntax structures. The parameter information necessary for the In addition, boundary condition correction information may be generated from information after demultiplexing as necessary. Thereafter, at least post-quantization information is notified to the inverse quantizer 805 and motion vector information is notified to the MC unit 811. Further, boundary information correction information may be notified to the boundary information determination unit 806 as necessary.

  The inverse quantizer 805 acquires post-quantization information from the entropy decoder 804, performs predetermined inverse quantization on the acquired post-quantization information (step S93), and performs a residual DCT after inverse quantization. Generates dequantized information that is coefficient information. The inverse quantizer 805 notifies the boundary condition determination unit 806 of the generated post-quantization information.

  Thereafter, it is determined whether or not intra decoding is performed (step S94). In the case of intra decoding, the switch 812 is switched to the terminal a side. That is, the switch 812 is disconnected so as not to acquire the predicted frame. Thereafter, the process proceeds to step S96.

  If it is not intra decoding, the MC unit 811 performs motion compensation from the motion vector information acquired from the entropy decoder 804 and the reference image frame acquired from the frame memory 810 (step S95). As a result, a predicted image frame is generated by the MC unit 811 and supplied to the terminal b side of the switch 812. Further, the switch 812 is switched to the terminal b side. Since the subsequent operation is the same as that of the embodiment shown in FIG. 3, the description thereof is omitted here. Hereinafter, an operation in the case of intra decoding will be described.

  The boundary condition determination unit 806 acquires boundary condition correction information from the entropy decoder 804, acquires post-quantization information from the inverse quantizer 805, and acquires the acquired boundary condition correction information and post-dequantization information. Based on the above, the boundary condition is determined (step S96). In this boundary condition determination, when boundary condition correction information can be acquired, the boundary condition determination unit 806 notifies the demultiplexed harmonic local orthogonal transformer 807 of the acquired boundary condition correction information as boundary condition determination information, and the boundary condition When the correction information cannot be acquired, the obtained post-quantization information is used to analyze the discontinuity of the image signal between the block to be decoded and the neighboring blocks adjacent thereto, and at least 1 It is determined whether the boundary condition needs to be corrected based on two predetermined determination conditions.

  Further, the boundary condition determiner 806 generates boundary condition correction information for specifying whether the boundary condition needs to be corrected, and notifies the demultiplex harmonic local orthogonal transformer 807 as boundary condition determination information. Usually, the boundary conditions between the blocks adjacent to the upper, lower, left, and right sides of the block currently being decoded are analyzed, and necessary boundary condition correction information is generated for each boundary. Further, the boundary condition determination unit 806 notifies the acquired dequantized information to the inverse multiple harmonic local orthogonal transformer 807.

  The inverse multiple harmonic local orthogonal transformer 807 performs a predetermined inverse multiple harmonic local orthogonal transform based on the boundary condition determination information and the dequantized information acquired from the boundary condition determiner 806 (step S97), and performs inverse quantization. Decoded DCT coefficient information obtained by decoding from the subsequent residual DCT coefficient information is generated. The inverse multiple harmonic local orthogonal transformer 807 notifies the inverse orthogonal transformer 808 of the generated decoded DCT coefficient information.

  The inverse orthogonal transformer 808 acquires the decoded DCT coefficient information from the inverse multiple harmonic local orthogonal transformer 807, performs inverse orthogonal transformation on the acquired decoded DCT coefficient information (step S98), generates a decoded image frame, In order to use the decoded image frame as a reference image frame, the frame memory 810 is notified through an adder 809.

  In the frame memory 810, an image frame is acquired and stored as a reference image frame. Further, a reference image frame is notified to the MC unit 811 as necessary. Here, the frame memory 810 preferably accumulates at least one reference image frame and can notify a necessary reference image frame in response to a request from each component. Also, the frame memory 810 outputs a decoded image frame in accordance with the display timing of the decoded image frame (step S99), and notifies an image output necessary for display as an output code 813 to an external output display.

  Thereafter, it is further determined whether there is an encoded bit stream to be decoded (step S100). When there is an encoded bit stream to be decoded, the process proceeds to step S91 to continue the decoding process, and when there is no need for decoding, a series of decoding processes are terminated.

  In the above description of the moving picture decoding apparatus shown in FIG. 25, the decoding process has been described as being performed in units of frames. However, the present invention is not limited to this, and frames such as macroblocks, blocks, and subblocks are used. You may comprise so that it may carry out by a smaller image unit. Further, a configuration may be adopted in which a plurality of macroblocks are performed in units of slices. Therefore, when decoding processing is next performed in units of macroblocks, the processing unit when performing inverse integer DCT that is inverse orthogonal transformation, inverse multiple harmonic local orthogonal transformation, or the like is performed in units of 4 × 4 sub-blocks. 23 will be described with reference to the flowchart of FIG.

  First, the input receiver 802 sequentially acquires predetermined packet information from a predetermined transmission path or a predetermined recording medium as an input code 801, and reconstructs an encoded bit stream. Thereafter, the demultiplexer 803 is notified of the encoded bit stream. The demultiplexer 803 acquires the encoded bitstream reconstructed by the input receiver 802, and demultiplexes the encoded information based on a predetermined syntax structure on the acquired encoded bitstream ( In step S91), the information after demultiplexing is notified to the entropy decoder 804.

  The entropy decoder 804 obtains information after demultiplexing and performs predetermined entropy decoding (step S92), whereby post-quantization information and motion vector information for configuring a decoding target block, And other parameter information necessary for constructing a predetermined syntax structure. In addition, boundary condition correction information may be generated from the information after demultiplexing as necessary. Thereafter, at least post-quantization information is notified to the inverse quantizer 805 and motion vector information is notified to the MC unit 811. Further, boundary information correction information may be notified to the boundary information determination unit 806 as necessary.

  The inverse quantizer 805 acquires post-quantization information of the block to be decoded from the entropy decoder 804. The obtained post-quantization information is subjected to predetermined de-quantization (step S93) to generate de-quantization information which is residual DCT coefficient information after the de-quantization, and the de-quantization information is defined as a boundary. The condition determination unit 806 is notified.

  Thereafter, it is determined whether or not intra decoding is performed (step S94). In the case of intra decoding, the switch 812 is switched to the terminal a side. That is, the switch 812 is disconnected so as not to acquire the predicted frame. Thereafter, the process proceeds to step S96. When it is not intra decoding, since it is the same as that of said description, description is abbreviate | omitted here. Hereinafter, an operation in the case of intra decoding will be described.

  Based on the boundary condition correction information acquired from the entropy decoder 804 and the post-inverse quantization information of the sub-block to be decoded acquired from the dequantizer 805, the boundary condition determiner 806 The condition is determined (step S96). If the boundary condition correction information can be acquired, the boundary condition determination unit 806 notifies the acquired boundary condition correction information to the inverse multiple harmonic local orthogonal transformer 807 as boundary condition determination information, and the boundary condition correction information is If it cannot be obtained, the obtained post-quantization information is used to analyze the discontinuity of the image signal between the sub-block to be decoded and the neighboring sub-blocks, and at least one It is determined whether the boundary condition needs to be corrected based on a predetermined determination condition.

  Further, the boundary condition determiner 806 generates boundary condition correction information for specifying whether the boundary condition needs to be corrected, and notifies the demultiplex harmonic local orthogonal transformer 807 as boundary condition determination information. Here, when intra decoding is performed in raster order in units of macroblocks, boundary conditions are analyzed only for adjacent sub-blocks that have been dequantized, and the boundary between adjacent sub-blocks before decoding is analyzed. As for the conditions, it is preferable to correct the boundary conditions without performing analysis. However, it is desirable to analyze boundary conditions for decoded sub-blocks within the same macroblock. The boundary condition determiner 806 notifies the acquired dequantized information to the demultiplexed harmonic local orthogonal transformer 807.

  The inverse multiple harmonic local orthogonal transformer 807 performs predetermined inverse multiple harmonic local orthogonal transformation based on the boundary condition determination information and the dequantized information acquired from the boundary condition determiner 806 (step S97), and performs inverse quantization. Decoded DCT coefficient information obtained by decoding from the subsequent residual DCT coefficient information is generated, and the generated decoded DCT coefficient information is notified to the inverse orthogonal transformer 808.

  The inverse orthogonal transformer 808 obtains the decoded DCT coefficient information from the inverse multiple harmonic local orthogonal transformer 807, performs inverse orthogonal transformation on the obtained decoded DCT coefficient information (step S98), and generates a decoded image block. The inverse orthogonal transformer 808 supplies the generated decoded image block to the frame memory 810 through the adder 809 so that the decoded image block can be used as a reference image frame.

  In the frame memory 810, the decoded image block is acquired and stored for use as a reference image frame. Here, the frame memory 810 preferably accumulates at least one reference image frame and can notify a necessary reference image frame in response to a request from each component. Also, the frame memory 810 outputs a decoded image frame in accordance with the display timing of the decoded image frame (step S99), and notifies an image output necessary for display as an output code 813 to an external output display.

  Thereafter, it is further determined whether there is an encoded bit stream to be decoded (step S100). When there is an encoded bit stream to be decoded, the process proceeds to step S91 to continue the decoding process, and when there is no need for decoding, a series of decoding processes are terminated.

  In the present embodiment, when decoding processing is performed in units of macroblocks, when analyzing boundary conditions between a subblock to be decoded and its adjacent and surrounding subblocks, in units of macroblocks When intra decoding is performed by raster order in, boundary conditions are analyzed only for the sub-quantized adjacent sub-blocks, and boundary conditions with the adjacent sub-blocks before decoding are analyzed. By correcting the boundary condition without performing it, it is possible to reduce the processing of the present embodiment and realize faster decoding.

  In the present invention, the functions of the moving picture coding apparatus and the moving picture decoding apparatus shown in FIGS. 23 and 25 are realized by the central processing control apparatuses 2303 and 2403 which are computers shown in FIGS. Includes programs. This program may be taken in from a recording medium or may be taken in via a communication network.

  FIG. 27 is a block diagram showing an example of an information processing apparatus that operates according to the encoding program of the present invention. In the figure, an information processing device 2300 operates in accordance with an embodiment of an input program 2301 for inputting various types of information, an output device 2302 for outputting various types of information, and an encoding program of the present invention. A central processing control device 2303, an external storage device 2304, a temporary storage device 2305 used as a work area for arithmetic processing by the central processing control device 2303, and a communication device 2306 for communicating with the outside are bidirectional. It is configured to be connected by a bus 2307.

The central processing control device 2303 receives the encoding program of the present invention from a recording medium or via a communication network and is taken in by the communication device 2306, and performs orthogonal transform means 2308 corresponding to the orthogonal transform unit 703 in FIG. Multiple harmonic local orthogonal transform means 2309 corresponding to the harmonic local orthogonal transformer 705, quantization means 2310 corresponding to the quantizer 706,
Inverse quantization means 2311 corresponding to the inverse quantizer 707, inverse multiple harmonic local orthogonal transform means 2312 equivalent to the inverse multiple harmonic local orthogonal transformer 709, inverse orthogonal transform means 2313 equivalent to the inverse orthogonal transformer 710, ME device ME means 2314 corresponding to 713, MC means 2315 equivalent to MC unit 714, entropy encoding means 2316 equivalent to entropy encoder 716, multiplexing means 2317 equivalent to multiplexer 717, and output transmitter 718 Output transmission means 2318, boundary condition analysis means 2319 corresponding to the boundary condition analyzer 704, and boundary condition determination means 2320 corresponding to the boundary condition determination unit 708, and the code of the present invention shown in FIG. The same operation as that of the computer is executed by software processing.

  FIG. 28 is a block diagram showing an example of an information processing apparatus that operates according to the decryption program of the present invention. In the figure, an information processing device 2400 includes an input device 2401 for inputting various kinds of information, an output device 2402 for outputting various kinds of information, and a central processing control device 2403 operated by the decryption program of the present invention. And an external storage device 2404, a temporary storage device 2405 used as a work area for arithmetic processing by the central processing control device 2403, and a communication device 2406 for communicating with the outside are connected by a bidirectional bus 2407. It is configured.

  The central processing control device 2403 receives the decoding program of the present invention from a recording medium or via a communication network and is taken in by the communication device 2306, and entropy decoding means 2408 corresponding to the entropy decoder 804 in FIG. , Inverse quantization means 2409 corresponding to inverse quantizer 805, inverse multiple harmonic local orthogonal transform means 2410 equivalent to inverse multiple harmonic local orthogonal transformer 807, inverse orthogonal transform means 2411 equivalent to inverse orthogonal transformer 808, MC MC means 2412 corresponding to the receiver 811, input receiving means 2413 corresponding to the input receiver 802, demultiplexing means 2414 corresponding to the demultiplexer 803, and boundary condition determining means 2415 corresponding to the boundary condition determiner 806. At least a function, and the same operation as that of the decoding device of the present invention shown in FIG. Executed by hardware processing.

It is a functional block diagram of one embodiment of an encoding device of the present invention. It is a flowchart for operation | movement description of one Embodiment of the encoding apparatus of this invention. It is a functional block diagram of one Embodiment of the decoding apparatus of this invention. It is a flowchart for operation | movement description of one Embodiment of the decoding apparatus of this invention. It is a functional block diagram of other embodiment of the encoding apparatus of this invention. It is a flowchart for operation | movement description of other embodiment of the encoding apparatus of this invention. It is a conceptual diagram for showing the positional relationship of each DCT coefficient information produced | generated by 4x4 DCT, and arrangement | positioning of a block. (3x3 block) It is a conceptual diagram of an example for showing a mode that the multiple harmonic local orthogonal transformation of a horizontal direction is performed with respect to the process target of FIG. It is a conceptual diagram of an example for showing a mode that the multiple harmonic local orthogonal transformation of a vertical direction is performed with respect to the process target of FIG. It is a conceptual diagram of the other example for showing a mode that the multiple harmonic local orthogonal transformation of the vertical direction is performed with respect to the process target of FIG. It is a conceptual diagram of the other example for showing a mode that horizontal multiple harmonic local orthogonal transformation is performed with respect to the process target of FIG. It is a conceptual diagram for showing the positional relationship of each DCT coefficient information produced | generated by 4x4 DCT, and arrangement | positioning of a block. (5x5 block) It is a conceptual diagram of an example for showing a mode that the multiple harmonic local orthogonal transformation of a horizontal direction is performed with respect to the process target of FIG. It is a conceptual diagram of an example for showing a mode that the multiple harmonic local orthogonal transformation of a vertical direction is performed with respect to the process target of FIG. It is a conceptual diagram of the other example for showing a mode that the multiple harmonic local orthogonal transformation of the vertical direction is performed with respect to the process target of FIG. It is a conceptual diagram of the other example for showing a mode that multi-harmonic local orthogonal transformation of a horizontal direction is performed with respect to the process target of FIG. It is a functional block diagram of other Embodiment of the decoding apparatus of this invention. It is a flowchart for operation | movement description of other embodiment of the decoding apparatus of this invention. It is a functional block diagram showing the basic composition of the information processor which realizes the function with which one embodiment of the coding device of the present invention is run by running the coding program of the present invention. It is a functional block diagram showing the basic composition of the information processor which realizes the function with which one embodiment of the decoding device of the present invention is executed by running the decoding program of the present invention. It is a functional block diagram showing the basic composition of the information processor which realizes the function with which other embodiments of the coding device of the present invention are run by running the coding program of the present invention. It is a functional block diagram showing the basic composition of the information processor which realizes the function with which other embodiments of the decoding device of the present invention are executed by running the decoding program of the present invention. It is a functional block diagram of 2nd Embodiment in the encoding apparatus of this invention. It is a flowchart for operation | movement description of 2nd Embodiment in the encoding apparatus of this invention. It is a functional block diagram of 2nd Embodiment in the decoding apparatus of this invention. It is a flowchart for operation | movement description of 2nd Embodiment in the decoding apparatus of this invention. It is a functional block diagram showing the basic composition of the information processor which realizes the function with which a 2nd embodiment is provided in the coding device of the present invention by running the coding program of the present invention. It is a functional block diagram showing the basic composition of the information processor which realizes the function with which a 2nd embodiment is provided in the decoding device of the present invention by running the decoding program of the present invention. It is a functional block diagram of an example (AVC) of the conventional encoding apparatus. It is a functional block diagram of an example (AVC) of the conventional decoding apparatus. It is a conceptual diagram for showing the basic concept of multiple harmonic local orthogonal transformation. It is a conceptual diagram for demonstrating the handling and influence of a signal at the time of performing normal DCT. It is a conceptual diagram for demonstrating the handling and influence of a signal at the time of performing multiple harmonic local orthogonal transformation. It is a conceptual diagram for showing the definition regarding each block and the boundary between blocks.

Explanation of symbols

101, 221, 701, 801 Input code 102 Subtractor 103, 703 Orthogonal transformer 104, 706 Quantizer 105, 225, 707, 805 Inverse quantizer 106, 204, 710, 808 Inverse orthogonal transformer 107, 205, 711, 809 Adder 108, 209 Deblocking filter 109, 210, 712, 810 Frame memory 110, 208 Intra predictor 111, 713 Motion estimator (ME) 112, 206, 714, 811 Motion compensator (MC) 113, 207, 715, 812 Switch 121, 705 Multiple harmonic local orthogonal transformer 122, 226, 709, 807 Inverse multiple harmonic local orthogonal transformer 123, 716 Entropy encoder 124, 717 Multiplexer 125, 719 Output transmitter 126, 211, 719, 813 Force code 131, 704 Boundary condition analyzer 132, 231, 708, 806 Boundary condition determiner 222, 802 Input receiver 223, 803 Demultiplexer 224, 804 Entropy decoder 1900, 2000, 2100, 2200, 2300, 2400 Information processing device 1901, 2001, 2011, 2201, 2301, 2401 Input device 1902, 2002, 2102, 2202, 2302, 2402 Output device 1903, 2003, 2103, 2203, 2303, 2403 Central processing control device 1904, 2004, 2104 2204, 2304, 2404 External storage device 1905, 2005, 2105, 2205, 2305, 2405 Temporary storage device 1906, 2006, 2106, 2206, 2306, 2406 Communication device

Claims (24)

A predicted image signal generated from a locally decoded image signal by subdividing a screen area of the input video signal to be encoded as a unit of a rectangular area having a predetermined number of pixels, and using the rectangular area as a processing unit; In an encoding device that generates a differential image signal that is a differential signal with the moving image signal to be encoded, and performs encoding on the differential image signal,
Orthogonal transform means for performing orthogonal transform on the difference image signal as a processing unit and outputting orthogonal transform coefficient information;
The encoding target satisfying the Poisson equation with the gradient of the moving image signal obtained from the orthogonal transform coefficient information at the boundary between the rectangular region adjacent to the encoding target and another rectangular region as a boundary condition; When generating approximate orthogonal transform coefficient information obtained by orthogonally transforming the estimated signal in the rectangular area, the approximate orthogonal transform coefficient information generated is approximate orthogonality obtained by inverse multiple harmonic local orthogonal transform performed in the local decoding process. In order to approximate the transform coefficient information, the orthogonal transform coefficient information obtained by performing inverse quantization after quantization on the orthogonal transform coefficient information acquired from the orthogonal transform unit is estimated based on a quantization parameter used in quantization. , the residual by calculating a difference between the estimation result to generate the approximate orthogonal transform coefficient information with the approximate orthogonal transform coefficient information generated as the orthogonal transform coefficient information And polyharmonic local orthogonal transform means for generating exchange transform coefficient information,
Quantization means for quantizing the residual orthogonal transform coefficient information based on a predetermined quantization parameter;
Entropy encoding means for generating predetermined post-entropy encoded information by performing entropy encoding on the residual orthogonal transform coefficient information after quantization output from the quantization means;
Multiplexing means for generating an encoded bitstream by performing predetermined multiplexing according to a predetermined syntax structure on the entropy-encoded information output from the entropy encoding means;
Inverse quantization means for inversely quantizing residual orthogonal transform coefficient information after quantization output from the quantization means based on a predetermined quantization parameter;
The Poisson equation that satisfies the boundary condition of the gradient of the moving image signal obtained from the residual orthogonal transform coefficient information after the inverse quantization at the boundary between the rectangular area to be decoded and another adjacent rectangular area is satisfied. In addition, the approximate orthogonal transform coefficient information obtained by orthogonal transform with respect to the estimation signal in the rectangular area to be decoded is the residual orthogonal transform coefficient information after the inverse quantization generated by the inverse quantization means. And demultiplexing the orthogonal transform coefficient information by combining the residual orthogonal transform coefficient information after dequantization and the generated approximate orthogonal transform coefficient information. Means,
A video encoding apparatus comprising: an inverse orthogonal transform unit configured to generate the local decoded image signal by performing an inverse orthogonal transform on the orthogonal transform coefficient information generated by the inverse multiple harmonic local orthogonal transform unit. . The orthogonal transform unit is a unit that uses an orthogonal transform base as a DCT base and generates DCT coefficient information as the orthogonal transform coefficient information.
The multi-harmonic local orthogonal transform means uses an orthogonal transform base as a DCT base, and calculates a slope of a moving image signal obtained from the DCT coefficient information at a boundary between the rectangular area to be encoded and another adjacent rectangular area. When generating approximate DCT coefficient information obtained by orthogonally transforming the estimated signal in the rectangular region to be encoded so as to satisfy the Poisson equation as a boundary condition , the generated approximate DCT coefficient information is locally In order to approximate the approximate DCT coefficient information obtained by the inverse multiple harmonic local orthogonal transform performed in the decoding process, the DCT coefficient information acquired from the orthogonal transform unit is quantized based on the quantization parameter used in the quantization. estimates the DCT coefficient information subjected to inverse quantization to generate the approximate DCT coefficient information by using this estimation result, obtained by said orthogonal transforming means A means for generating the residual DCT coefficient information by taking the difference between the approximate DCT coefficient information which has been generated and DCT coefficient information is,
The inverse multiple harmonic local orthogonal transform means uses an orthogonal transform base as a DCT base, and calculates a slope of a moving image signal obtained from the DCT coefficient information at a boundary between the rectangular area to be decoded and another adjacent rectangular area. Approximate DCT coefficient information obtained by orthogonal transform with respect to the estimated signal in the rectangular area to be decoded that satisfies the Poisson equation as a boundary condition is obtained by inverse quantization by the inverse quantization means. The residual DCT coefficient information after the inverse quantization obtained by the inverse quantization and the residual DCT coefficient information after the inverse quantization obtained by the inverse quantization and the generated approximate DCT coefficient information. The moving picture coding apparatus according to claim 1 , wherein the moving picture coding apparatus is means for generating the DCT coefficient information by performing synthesis. Based on the orthogonal transform coefficient information generated by the orthogonal transform unit, analyze the discontinuity of the video signal between the rectangular region to be encoded and the adjacent rectangular region, First boundary condition correction information for determining whether or not boundary condition correction is necessary based on at least one predetermined determination condition and for specifying whether or not boundary condition correction is necessary A boundary condition analysis means to generate;
If the first boundary condition correction information is not valid, the surrounding area adjacent to the rectangular area to be decoded based on the residual orthogonal transform coefficient information after inverse quantization by the inverse quantization means Analyzing the discontinuity of the moving image signal with respect to the rectangular region of the image, and determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition. Second boundary condition correction information for specifying whether or not it is necessary is generated, and the generated second boundary condition correction information or the valid first boundary condition correction information is used as boundary condition determination information. Boundary condition judging means for
Further comprising
The multiple harmonic local orthogonal transform means obtains the inclination of the moving image signal obtained from the orthogonal transform coefficient information which is a boundary condition of each rectangular region based on the first boundary condition correction information obtained by the boundary condition analysis means. Approximate orthogonal transform coefficient information that is orthogonally transformed with respect to the estimated signal in the rectangular region that is to be encoded so as to satisfy the Poisson equation based on the boundary condition of each rectangular region The residual orthogonal transform coefficient information is generated by taking the difference between the orthogonal transform coefficient information obtained by the orthogonal transform and the generated approximate orthogonal transform coefficient information. Means to generate,
The inverse multiple harmonic local orthogonal transform means identifies the discontinuity of the boundary portion of the rectangular region from the boundary condition determination information acquired from the boundary condition determination means, and when the target boundary portion is discontinuous Correcting the inclination of the moving image signal obtained from the residual orthogonal transform coefficient information, and the residual orthogonality after the inverse quantization at the boundary of the rectangular area to be decoded with the adjacent rectangular area Approximate orthogonal transformation coefficient information obtained by orthogonal transformation with respect to the estimation signal in the rectangular region to be decoded, which satisfies the Poisson equation with the gradient of the moving image signal obtained from the transformation coefficient information as a boundary condition, The residual orthogonal transform coefficient information after the inverse quantization generated by the inverse quantization means is used to generate the residual orthogonal transform coefficient information after the inverse quantization and the generated approximate orthogonal transform coefficient information. Moving picture coding apparatus according to claim 1, characterized in that the means for decoding the orthogonal transform coefficient information by performing the synthesis of the. In an encoding device that subdivides a screen area of an input video signal to be encoded as a unit of a rectangular area having a predetermined number of pixels and performs intra encoding using the rectangular area as a processing unit.
Orthogonal transform means for performing orthogonal transform using the rectangular area as a processing unit, and outputting orthogonal transform coefficient information for each rectangular area in advance over the entire screen area of the moving image signal to be encoded;
In order to analyze the discontinuity of the moving image signal existing in one screen region of the moving image signal to be encoded, the rectangle existing around the rectangular region to be encoded Analyzing the discontinuity of the moving image signal between the regions and determining whether or not the boundary condition needs to be corrected based on a predetermined determination condition, and whether or not the boundary condition needs to be corrected Boundary condition analysis means for generating first boundary condition correction information for specifying
The encoding target satisfying the Poisson equation with the gradient of the moving image signal obtained from the orthogonal transform coefficient information at the boundary between the rectangular region adjacent to the encoding target and another rectangular region as a boundary condition; When generating approximate orthogonal transform coefficient information obtained by orthogonally transforming the estimated signal in the rectangular area, the approximate orthogonal transform coefficient information generated is approximate orthogonality obtained by inverse multiple harmonic local orthogonal transform performed in the local decoding process. In order to approximate the transform coefficient information, the orthogonal transform coefficient information obtained by performing inverse quantization after quantization on the orthogonal transform coefficient information acquired from the orthogonal transform unit is estimated based on a quantization parameter used in quantization. , the residual by calculating a difference between the estimation result to generate the approximate orthogonal transform coefficient information with the approximate orthogonal transform coefficient information generated as the orthogonal transform coefficient information And polyharmonic local orthogonal transform means for generating exchange transform coefficient information,
Quantization means for quantizing the residual orthogonal transform coefficient information based on a predetermined quantization parameter;
Entropy encoding means for generating predetermined post-entropy encoded information by performing entropy encoding on the residual orthogonal transform coefficient information after quantization output from the quantization means;
Multiplexing means for generating an encoded bitstream by performing predetermined multiplexing according to a predetermined syntax structure on the entropy-encoded information output from the entropy encoding means;
Inverse quantization means for inversely quantizing residual orthogonal transform coefficient information after quantization output from the quantization means based on a predetermined quantization parameter;
If the first boundary condition correction information is not valid, the surrounding area adjacent to the rectangular area to be decoded based on the residual orthogonal transform coefficient information after inverse quantization by the inverse quantization means Analyzing the discontinuity of the moving image signal with respect to the rectangular region of the image, and determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition. Second boundary condition correction information for specifying whether or not it is necessary is generated, and the generated second boundary condition correction information or the valid first boundary condition correction information is used as boundary condition determination information. Boundary condition judging means for
The Poisson equation that satisfies the boundary condition of the gradient of the moving image signal obtained from the residual orthogonal transform coefficient information after the inverse quantization at the boundary between the rectangular area to be decoded and another adjacent rectangular area is satisfied. In addition, the approximate orthogonal transform coefficient information obtained by orthogonal transform with respect to the estimation signal in the rectangular area to be decoded is the residual orthogonal transform coefficient information after the inverse quantization generated by the inverse quantization means. And demultiplexing the orthogonal transform coefficient information by combining the residual orthogonal transform coefficient information after dequantization and the generated approximate orthogonal transform coefficient information. Means,
An inverse orthogonal transform unit that generates the local decoded image signal by performing an inverse orthogonal transform on the orthogonal transform coefficient information generated by the inverse multiple harmonic local orthogonal transform unit;
And a moving picture coding apparatus that realizes intra coding. The orthogonal transformation means is means for performing orthogonal transformation with the rectangular area as a processing unit and outputting the generated orthogonal transformation coefficient information in the processing unit of the rectangular area, and the boundary condition analysis means is an encoding target. Analyzing the discontinuity existing in the moving image signal, the discontinuity between the rectangular area to be processed and a predetermined rectangular area that has been orthogonally transformed, First boundary condition correction information for determining whether or not boundary condition correction is necessary based on at least one predetermined determination condition and for specifying whether or not boundary condition correction is necessary The quantization means quantizes the residual orthogonal transform coefficient information based on the predetermined quantization parameter, and outputs the generated post-quantization information in units of processing of the rectangular area. Means, The inverse quantization means performs inverse quantization on the quantized residual orthogonal transform coefficient information output from the quantization means based on the predetermined quantization parameter, and generates the generated inverse quantization after the inverse quantization Means for outputting residual orthogonal transform coefficient information in units of processing of the rectangular area, and the boundary condition determination means is generated by the inverse quantization means when the first boundary condition correction information is not valid. Based on the residual orthogonal transform coefficient information after dequantization, analyzing the discontinuity of the video signal between the rectangular area to be decoded and the adjacent rectangular area, Second boundary condition correction information for determining whether or not boundary condition correction is necessary based on at least one predetermined determination condition and for specifying whether or not boundary condition correction is necessary Generated and the generated second A means for the boundary condition determining information the first boundary condition correction information is field condition correction information or effective, according to claim 4, wherein the encoding is performed as a processing unit prescribed the rectangular region Video encoding device. The orthogonal transform unit generates DCT coefficient information as the orthogonal transform coefficient information by performing orthogonal transform using a DCT base as an orthogonal transform base, and the boundary condition analyzing unit is configured to generate the DCT coefficient generated by the orthogonal transform unit. Based on the coefficient information, the discontinuity of the moving image signal between the rectangular area to be encoded and the adjacent rectangular area is analyzed, and the boundary condition is based on at least one predetermined determination condition. Generating the first boundary condition correction information for specifying whether correction of the boundary condition is necessary, and determining whether correction of the first boundary condition is necessary,
The multiple harmonic local orthogonal transform means sets the orthogonal transform base as a DCT base, and controls the inclination of the moving image signal obtained from the DCT coefficient information which is a boundary condition of each rectangular region based on the first boundary condition correction information. And an encoding target that satisfies the Poisson's equation with the gradient of the moving image signal obtained from the DCT coefficient information at the boundary of the rectangular region as the encoding target adjacent to the other rectangular region as a boundary condition The approximate DCT coefficient information obtained by performing orthogonal transform on the estimated signal in the rectangular area is generated using the DCT coefficient information after the orthogonal transform, and then the DCT coefficient information after the orthogonal transform. And generating the residual DCT coefficient information by taking a difference between the generated approximate DCT coefficient information and
When the first boundary condition correction information is not valid, the boundary condition determination unit determines whether the rectangular area to be decoded based on the residual DCT coefficient information and an adjacent rectangular area are adjacent to each other. The discontinuity of the moving image signal is analyzed to determine whether the boundary condition needs to be corrected based on at least one predetermined determination condition, and whether the boundary condition needs to be corrected Means for generating second boundary condition correction information for specifying, and using the second boundary condition correction information or the effective first boundary condition correction information as the boundary condition determination information;
The inverse multiple harmonic local orthogonal transform means uses the orthogonal transform base as a DCT base, specifies the discontinuity of the boundary portion of the rectangular area from the boundary condition determination information acquired from the boundary condition determination means, Is discontinuous, the inclination of the moving image signal obtained from the residual DCT coefficient information is corrected, and the inverse quantum at the boundary of the rectangular region to be decoded is adjacent to another rectangular region. Approximate DCT obtained by orthogonal transform with respect to the estimation signal in the rectangular region to be decoded so as to satisfy the Poisson equation having the boundary condition of the gradient of the moving image signal obtained from the residual DCT coefficient information after conversion. The coefficient information is generated by using the residual DCT coefficient information after the inverse quantization generated by the inverse quantization means, the residual DCT coefficient information after the inverse quantization, and the generated Moving picture coding apparatus as claimed in any one of claims 3 to 5 that, wherein a means for decoding the DCT coefficient information by performing the synthesis of the similar DCT coefficient information. A predicted image signal generated from a locally decoded image signal by subdividing a screen area of the input video signal to be encoded as a unit of a rectangular area having a predetermined number of pixels, and using the rectangular area as a processing unit; In a moving image encoding program for causing a computer to generate a difference image signal that is a difference signal from the moving image signal to be encoded, and to perform encoding on the difference image signal,
The computer,
Orthogonal transform means for performing orthogonal transform on the difference image signal as a processing unit and outputting orthogonal transform coefficient information;
The encoding target satisfying the Poisson equation with the gradient of the moving image signal obtained from the orthogonal transform coefficient information at the boundary between the rectangular region adjacent to the encoding target and another rectangular region as a boundary condition; When generating approximate orthogonal transform coefficient information obtained by orthogonally transforming the estimated signal in the rectangular area, the approximate orthogonal transform coefficient information generated is approximate orthogonality obtained by inverse multiple harmonic local orthogonal transform performed in the local decoding process. In order to approximate the transform coefficient information, the orthogonal transform coefficient information obtained by performing inverse quantization after quantization on the orthogonal transform coefficient information acquired from the orthogonal transform unit is estimated based on a quantization parameter used in quantization. , the residual by calculating a difference between the estimation result to generate the approximate orthogonal transform coefficient information with the approximate orthogonal transform coefficient information generated as the orthogonal transform coefficient information And polyharmonic local orthogonal transform means for generating exchange transform coefficient information,
Quantization means for quantizing the residual orthogonal transform coefficient information based on a predetermined quantization parameter;
Entropy encoding means for generating predetermined post-entropy encoded information by performing entropy encoding on the residual orthogonal transform coefficient information after quantization output from the quantization means;
Multiplexing means for generating an encoded bitstream by performing predetermined multiplexing according to a predetermined syntax structure on the entropy-encoded information output from the entropy encoding means;
Inverse quantization means for inversely quantizing residual orthogonal transform coefficient information after quantization output from the quantization means based on a predetermined quantization parameter;
The Poisson equation that satisfies the boundary condition of the gradient of the moving image signal obtained from the residual orthogonal transform coefficient information after the inverse quantization at the boundary between the rectangular area to be decoded and another adjacent rectangular area is satisfied. In addition, the approximate orthogonal transform coefficient information obtained by orthogonal transform with respect to the estimation signal in the rectangular area to be decoded is the residual orthogonal transform coefficient information after the inverse quantization generated by the inverse quantization means. And demultiplexing the orthogonal transform coefficient information by combining the residual orthogonal transform coefficient information after dequantization and the generated approximate orthogonal transform coefficient information. Means,
A moving image code that functions as an inverse orthogonal transform unit that generates the local decoded image signal by performing an inverse orthogonal transform on the orthogonal transform coefficient information generated by the inverse multiple harmonic local orthogonal transform unit Program. The computer,
The orthogonal transform means functions as a means for generating DCT coefficient information as the orthogonal transform coefficient information, with the orthogonal transform base as a DCT base,
The multi-harmonic local orthogonal transform means uses an orthogonal transform base as a DCT base, and sets a gradient of a moving image signal obtained from the DCT coefficient information at a boundary between the rectangular area to be encoded and another adjacent rectangular area as a boundary. When generating approximate DCT coefficient information obtained by orthogonally transforming the estimated signal in the rectangular area to be encoded so as to satisfy the Poisson equation as a condition , the generated approximate DCT coefficient information is locally decoded. In order to approximate the approximate DCT coefficient information obtained by the inverse multiharmonic local orthogonal transform performed in the process, the DCT coefficient information acquired from the orthogonal transform unit is inverted after quantization based on the quantization parameter used in quantization. estimates the DCT coefficient information subjected to quantization to generate the approximate DCT coefficient information by using this estimation result, obtained et by the orthogonal transform means And it is made to function as means for generating the residual DCT coefficient information in taking the difference between the DCT coefficient information generated the approximate DCT coefficient information,
The inverse multiple harmonic local orthogonal transform means uses an orthogonal transform base as a DCT base, and calculates a slope of a moving image signal obtained from the DCT coefficient information at a boundary between the rectangular area to be decoded and another adjacent rectangular area. Approximate DCT coefficient information obtained by orthogonal transform with respect to the estimated signal in the rectangular area to be decoded that satisfies the Poisson equation as a boundary condition is obtained by inverse quantization by the inverse quantization means. The residual DCT coefficient information after the inverse quantization obtained by the inverse quantization and the residual DCT coefficient information after the inverse quantization obtained by the inverse quantization and the generated approximate DCT coefficient information. 8. The moving picture coding program according to claim 7 , wherein the moving picture coding program is caused to function as means for generating the DCT coefficient information by combining. The computer,
Based on the orthogonal transform coefficient information generated by the orthogonal transform unit, analyze the discontinuity of the video signal between the rectangular region to be encoded and the adjacent rectangular region, First boundary condition correction information for determining whether or not boundary condition correction is necessary based on at least one predetermined determination condition and for specifying whether or not boundary condition correction is necessary A boundary condition analysis means to generate;
If the first boundary condition correction information is not valid, the surrounding area adjacent to the rectangular area to be decoded based on the residual orthogonal transform coefficient information after inverse quantization by the inverse quantization means Analyzing the discontinuity of the moving image signal with respect to the rectangular region of the image, and determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition. Boundary that generates second boundary condition correction information for specifying whether or not it is necessary and uses the second boundary condition correction information or the effective first boundary condition correction information as boundary condition determination information Condition judging means;
To further function,
The multiple harmonic local orthogonal transform means controls the inclination of the moving image signal which is a boundary condition of each rectangular region based on the first boundary condition correction information obtained by the boundary condition analysis means, and Approximate orthogonal transform coefficient information obtained by orthogonal transform with respect to the estimated signal in the rectangular region that is the encoding target satisfying the Poisson equation based on the boundary condition of the rectangular region is the orthogonal transform after the orthogonal transform. It is generated using coefficient information and functions as means for generating the residual orthogonal transform coefficient information by taking the difference between the orthogonal transform coefficient information obtained by orthogonal transform and the generated approximate orthogonal transform coefficient information. ,
The inverse multiple harmonic local orthogonal transform means identifies the discontinuity of the boundary portion of the rectangular region from the boundary condition determination information acquired from the boundary condition determination means, and when the target boundary portion is discontinuous Poisson which corrects the inclination of the moving image signal and uses the inclination of the moving image signal obtained from the orthogonal transform coefficient information at the boundary of the rectangular area to be decoded as a boundary condition as a boundary condition The approximate orthogonal transform coefficient information obtained by orthogonal transform with respect to the estimated signal in the rectangular region to be decoded that satisfies the equation, the residual orthogonal transform coefficient information after dequantization, and the boundary Generated using the boundary condition determination information obtained by the condition determination means, and synthesizes the residual orthogonal transform coefficient information after inverse quantization and the generated approximate orthogonal transform coefficient information According to claim 7, wherein the moving picture encoding program for causing to function as means for generating the orthogonal transform coefficient information with the. One screen area of the input video signal to be encoded is subdivided with a rectangular area having a predetermined number of pixels as a unit, and intra encoding is performed with the rectangular area as a processing unit. In a video encoding program for
The computer,
Orthogonal transform means for performing orthogonal transform using the rectangular area as a processing unit, and outputting orthogonal transform coefficient information for each rectangular area in advance over the entire screen area of the moving image signal to be encoded;
In order to analyze the discontinuity of the moving image signal existing in one screen region of the moving image signal to be encoded, the rectangle existing around the rectangular region to be encoded Analyzing the discontinuity of the moving image signal between the regions and determining whether or not the boundary condition needs to be corrected based on a predetermined determination condition, and whether or not the boundary condition needs to be corrected Boundary condition analysis means for generating first boundary condition correction information for specifying
The encoding target satisfying the Poisson equation with the gradient of the moving image signal obtained from the orthogonal transform coefficient information at the boundary between the rectangular region adjacent to the encoding target and another rectangular region as a boundary condition; When generating approximate orthogonal transform coefficient information obtained by orthogonally transforming the estimated signal in the rectangular area, the approximate orthogonal transform coefficient information generated is approximate orthogonality obtained by inverse multiple harmonic local orthogonal transform performed in the local decoding process. In order to approximate the transform coefficient information, the orthogonal transform coefficient information obtained by performing inverse quantization after quantization on the orthogonal transform coefficient information acquired from the orthogonal transform unit is estimated based on a quantization parameter used in quantization. , the residual by calculating a difference between the estimation result to generate the approximate orthogonal transform coefficient information with the approximate orthogonal transform coefficient information generated as the orthogonal transform coefficient information And polyharmonic local orthogonal transform means for generating exchange transform coefficient information,
Quantization means for quantizing the residual orthogonal transform coefficient information based on a predetermined quantization parameter;
Entropy encoding means for generating predetermined post-entropy encoded information by performing entropy encoding on the residual orthogonal transform coefficient information after quantization output from the quantization means;
Multiplexing means for generating an encoded bitstream by performing predetermined multiplexing according to a predetermined syntax structure on the entropy-encoded information output from the entropy encoding means;
Inverse quantization means for inversely quantizing residual orthogonal transform coefficient information after quantization output from the quantization means based on a predetermined quantization parameter;
If the first boundary condition correction information is not valid, the surrounding area adjacent to the rectangular area to be decoded based on the residual orthogonal transform coefficient information after inverse quantization by the inverse quantization means Analyzing the discontinuity of the moving image signal with respect to the rectangular region of the image, and determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition. Second boundary condition correction information for specifying whether or not it is necessary is generated, and the generated second boundary condition correction information or the valid first boundary condition correction information is used as boundary condition determination information. Boundary condition judging means for
The Poisson equation that satisfies the boundary condition of the gradient of the moving image signal obtained from the residual orthogonal transform coefficient information after the inverse quantization at the boundary between the rectangular area to be decoded and another adjacent rectangular area is satisfied. In addition, the approximate orthogonal transform coefficient information obtained by orthogonal transform with respect to the estimation signal in the rectangular area to be decoded is the residual orthogonal transform coefficient information after the inverse quantization generated by the inverse quantization means. And demultiplexing the orthogonal transform coefficient information by combining the residual orthogonal transform coefficient information after dequantization and the generated approximate orthogonal transform coefficient information. Means,
A moving image code that functions as an inverse orthogonal transform unit that generates the local decoded image signal by performing an inverse orthogonal transform on the orthogonal transform coefficient information generated by the inverse multiple harmonic local orthogonal transform unit Program. The computer,
The orthogonal transform means performs orthogonal transform with the rectangular area as a processing unit, and functions as means for outputting the generated orthogonal transform coefficient information in the processing unit of the rectangular area. When analyzing the discontinuity existing in the moving image signal, the discontinuity between the rectangular area to be processed and the predetermined rectangular area that has been orthogonally transformed is analyzed. First boundary condition correction information for determining whether or not the boundary condition needs to be corrected based on at least one predetermined determination condition and for specifying whether or not the boundary condition needs to be corrected The quantization means quantizes the residual orthogonal transform coefficient information based on the predetermined quantization parameter, and the generated post-quantization information is processed in units of the rectangular area. output The inverse quantization means performs the inverse quantization on the quantized residual orthogonal transform coefficient information output from the quantization means based on the predetermined quantization parameter, and generates The residual orthogonal transform coefficient information after inverse quantization is functioned as a means for outputting in units of processing of the rectangular area, and the boundary condition determination means is configured such that when the first boundary condition correction information is not valid, Based on the residual orthogonal transform coefficient information after the inverse quantization generated by the inverse quantization means, the video signal between the rectangular area to be decoded and the adjacent rectangular area is decoded. A first analysis is performed for analyzing the discontinuity, determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition, and specifying whether the boundary condition needs to be corrected. 2 boundary condition correction By generating the information and functioning the generated second boundary condition correction information or the valid first boundary condition correction information as boundary condition determination information, the predetermined rectangular area is used as a processing unit. The moving image encoding program according to claim 10, wherein encoding is performed. The computer,
The orthogonal transform means functions to generate DCT coefficient information as the orthogonal transform coefficient information by performing orthogonal transform using a DCT base as an orthogonal transform base, and the boundary condition analysis means is generated by the orthogonal transform means Based on the DCT coefficient information, the video signal discontinuity between the rectangular area to be encoded and the adjacent rectangular area is analyzed, and at least one predetermined determination condition is satisfied. Determining whether or not the boundary condition needs to be corrected based on the function and generating the first boundary condition correction information for specifying whether or not the boundary condition needs to be corrected;
The multiple harmonic local orthogonal transform means sets the orthogonal transform base as a DCT base, and controls the inclination of the moving image signal obtained from the DCT coefficient information which is a boundary condition of each rectangular region based on the first boundary condition correction information. And an encoding target that satisfies the Poisson's equation with the gradient of the moving image signal obtained from the DCT coefficient information at the boundary of the rectangular region as the encoding target adjacent to the other rectangular region as a boundary condition The approximate DCT coefficient information obtained by performing orthogonal transform on the estimated signal in the rectangular area is generated using the DCT coefficient information after the orthogonal transform, and then the DCT coefficient information after the orthogonal transform. And a function of generating the residual DCT coefficient information by taking a difference between the generated approximate DCT coefficient information and
When the first boundary condition correction information is not valid, the boundary condition determination unit determines whether the rectangular area to be decoded based on the residual DCT coefficient information and an adjacent rectangular area are adjacent to each other. The discontinuity of the moving image signal is analyzed to determine whether the boundary condition needs to be corrected based on at least one predetermined determination condition, and whether the boundary condition needs to be corrected Generating second boundary condition correction information for identification, and functioning the second boundary condition correction information or the effective first boundary condition correction information as the boundary condition determination information;
The inverse multiple harmonic local orthogonal transform means uses the orthogonal transform base as a DCT base, specifies the discontinuity of the boundary portion of the rectangular area from the boundary condition determination information acquired from the boundary condition determination means, Is discontinuous, the inclination of the moving image signal obtained from the residual DCT coefficient information is corrected, and the inverse quantization at the boundary of the rectangular area to be decoded is adjacent to the other rectangular area Approximation obtained by orthogonal transform with respect to the estimated signal in the rectangular area to be decoded so as to satisfy the Poisson equation with the gradient of the moving image signal obtained from the residual DCT coefficient information as a boundary condition. The residual DCT coefficient information after the inverse quantization obtained by inverse quantization of the DCT coefficient information by the inverse quantization means and the boundary obtained by the boundary condition determination means After generating using the condition determination information, functioning as a means for generating the DCT coefficient information by combining the residual DCT coefficient information after dequantization and the generated approximate orthogonal transform coefficient information 12. The moving picture encoding program according to claim 9 , wherein the moving picture encoding program is any one of claims 9 to 11 . A coded bit stream generated by the moving picture coding apparatus according to claim 1 or 2 or the information processing apparatus operated by the moving picture coding program according to claim 7 or 8 is stored in a predetermined storage medium or a predetermined transmission path. A video decoding device that performs a decoding operation on the acquired encoded bitstream and outputs a decoded video signal,
Demultiplexing means for demultiplexing encoded information based on a predetermined syntax structure for the encoded bitstream;
Predemultiplexed information obtained from the demultiplexing means is subjected to predetermined entropy decoding, and post-quantization information, motion vector information, and parameter information necessary to construct a predetermined syntax structure Entropy decoding means to generate;
Inverse quantization means for generating residual DCT coefficient information as the residual orthogonal transform coefficient information after inverse quantization by performing inverse quantization on the quantized information acquired from the entropy decoding means; ,
A Poisson equation using a DCT basis as an orthogonal transformation basis and a gradient condition of a moving image signal obtained from the orthogonal transformation coefficient information at a boundary between the rectangular region to be decoded and another adjacent rectangular region as a boundary condition After generating the approximate DCT coefficient information orthogonally transformed with respect to the estimation signal in the rectangular region that is to be decoded, satisfying using the residual DCT coefficient information after the inverse quantization, An inverse multiple harmonic local orthogonal transform means for generating decoded DCT coefficient information by combining the residual DCT coefficient information obtained by inverse quantization and the generated approximate DCT coefficient information ;
An inverse orthogonal transform unit that generates a decoded differential image signal for one screen by performing an inverse orthogonal transform on the decoded DCT coefficient information acquired from the demultiplex harmonic local orthogonal transform unit;
A predicted moving image obtained by synthesizing a predicted image signal generated by predetermined motion compensation using the motion vector information acquired from the entropy decoding unit or a predicted image signal generated by intra prediction and the decoded difference image signal. And a video decoding device for generating a signal. An encoded bit stream generated by the moving image encoding apparatus according to claim 3 or 6 or the information processing apparatus operated by the moving image encoding program according to claim 9 or 12 is stored in a predetermined storage medium or a predetermined transmission path. A video decoding device that performs a decoding operation on the acquired encoded bitstream and outputs a decoded video signal,
Demultiplexing means for demultiplexing encoded information based on a predetermined syntax structure for the encoded bitstream;
Predemultiplexed information obtained from the demultiplexing means is subjected to predetermined entropy decoding to generate post-quantization information, motion vector information, and parameter information necessary for constructing a predetermined syntax structure. Entropy decoding means for extracting the boundary condition correction information encoded and incorporated, and
Inverse quantization means for generating the residual orthogonal transform coefficient information after inverse quantization by performing inverse quantization on the post-quantization information obtained from the entropy decoding means;
If the boundary condition correction information acquired from the entropy decoding means is not valid, the rectangle to be decoded is based on the post-quantization residual orthogonal transform coefficient information acquired from the inverse quantization means. Analyzing the discontinuity of the moving image signal between the region and the adjacent surrounding rectangular region, determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition, Generate boundary condition correction information for specifying whether or not condition correction is necessary, and use the generated boundary condition correction information or valid boundary condition correction information acquired from the entropy decoding means as a boundary condition Boundary condition determination means as determination information;
The boundary condition determination information and the residual orthogonal transform coefficient information after the dequantization are acquired, and the gradient of the moving image signal at the boundary between the rectangular area to be decoded and another adjacent rectangular area is defined as a boundary. The approximate orthogonal transform coefficient information obtained by performing orthogonal transform on the estimated signal in the rectangular area to be decoded that satisfies the Poisson equation as a condition is the residual orthogonal transform coefficient after the inverse quantization. Inverse multiple harmonic local orthogonality that generates decoded orthogonal transform coefficient information by combining the residual orthogonal transform coefficient information after dequantization and the generated approximate orthogonal transform coefficient information after generating using information Conversion means;
An inverse orthogonal transform unit that generates a decoded difference image signal for one screen by performing an inverse orthogonal transform on the decoded orthogonal transform coefficient information acquired from the demultiplex harmonic local orthogonal transform unit;
A predicted moving image obtained by synthesizing a predicted image signal generated by predetermined motion compensation using the motion vector information acquired from the entropy decoding unit or a predicted image signal generated by intra prediction and the decoded difference image signal. And a video decoding device for generating a signal. The residual orthogonal transform coefficient information after dequantization obtained by the dequantization means is residual DCT coefficient information, and the boundary condition determination means determines that the residual when the boundary condition correction information is not valid. Based on the DCT coefficient information, the image signal discontinuity between the rectangular area to be decoded and the adjacent rectangular area is analyzed, and the boundary condition is corrected based on at least one predetermined determination condition. Means for generating boundary condition correction information for specifying whether or not boundary condition correction is necessary, and using the obtained boundary condition correction information as boundary condition determination information The inverse multiple harmonic local orthogonal transform means uses a DCT base as an orthogonal transform base, and performs orthogonal transform on the estimated signal in the rectangular region to be decoded so as to satisfy the Poisson equation. The generated approximate DCT coefficient information is generated using the residual DCT coefficient information after the inverse quantization, and then the residual DCT coefficient information obtained by the inverse quantization and the generated approximate DCT coefficient information are combined. 15. The moving picture decoding apparatus according to claim 14 , wherein the moving picture decoding apparatus is means for generating decoded DCT coefficient information by performing. Obtaining a coded bitstream generated by the moving picture coding apparatus according to claim 4 or the information processing apparatus operated by the moving picture coding program according to claim 10 from a predetermined storage medium or a predetermined transmission path; A video decoding device that performs a decoding operation on the acquired encoded bitstream and outputs a decoded video signal,
Demultiplexing means for demultiplexing encoded information based on a predetermined syntax structure for the encoded bitstream;
The demultiplexed information obtained from the demultiplexing means is subjected to predetermined entropy decoding to generate post-quantization information and parameter information necessary for constructing a predetermined syntax structure, Entropy decoding means for extracting the boundary condition correction information incorporated into
Inverse quantization means for generating the residual orthogonal transform coefficient information after inverse quantization by performing inverse quantization on the post-quantization information obtained from the entropy decoding means;
If the boundary condition correction information acquired from the entropy decoding means is not valid, the rectangle to be decoded is based on the post-quantization residual orthogonal transform coefficient information acquired from the inverse quantization means. Analyzing the discontinuity of the moving image signal between the region and the adjacent surrounding rectangular region, determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition, Generate boundary condition correction information for specifying whether or not condition correction is necessary, and use the generated boundary condition correction information or valid boundary condition correction information acquired from the entropy decoding means as a boundary condition Boundary condition determination means as determination information;
The boundary condition determination information and the residual orthogonal transform coefficient information after the inverse quantization are acquired, and obtained from the orthogonal transform coefficient information at a boundary between the rectangular area to be decoded and another adjacent rectangular area. The approximate orthogonal transform coefficient information obtained by orthogonal transform with respect to the estimation signal in the rectangular region to be decoded, which satisfies the Poisson equation with the gradient of the moving image signal as a boundary condition, is subjected to the inverse quantization. After generating using the residual orthogonal transform coefficient information, the decoded orthogonal transform coefficient information is obtained by combining the residual orthogonal transform coefficient information after the inverse quantization and the generated approximate orthogonal transform coefficient information. Generating inverse multiple harmonic local orthogonal transform means;
An inverse orthogonal transform unit that generates a decoded image signal for one screen by performing an inverse orthogonal transform on the decoded orthogonal transform coefficient information acquired from the inverse multiple harmonic local orthogonal transform unit. Image decoding device. An encoded bitstream generated by the moving image encoding device according to claim 5 or the information processing device operated by the moving image encoding program according to claim 11 is acquired from a predetermined storage medium or a predetermined transmission path, A video decoding device that performs a decoding operation on the acquired encoded bitstream and outputs a decoded video signal,
Demultiplexing means for demultiplexing encoded information based on a predetermined syntax structure for the encoded bitstream;
The demultiplexed information obtained from the demultiplexing means is subjected to predetermined entropy decoding to generate post-quantization information and parameter information necessary for constructing a predetermined syntax structure, Entropy decoding means for extracting the boundary condition correction information incorporated into
By performing inverse quantization on the post-quantization information acquired from the entropy decoding means using a predetermined rectangular area as a processing unit, the residual orthogonal transform coefficient information after the inverse quantization is generated, and Dequantization means for outputting residual orthogonal transform coefficient information after dequantization in units of processing of the rectangular region;
When the boundary condition correction information acquired from the entropy decoding means is not valid, the boundary condition correction information is a decoding target based on the residual orthogonal transform coefficient information after the inverse quantization acquired from the inverse quantization means. Analyzing the discontinuity of the moving image signal between the rectangular area and the adjacent rectangular area, and determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition , Generating boundary condition correction information for specifying whether or not boundary condition correction is necessary, and generating the boundary condition correction information generated or the effective boundary condition correction information acquired from the entropy decoding means Boundary condition determination means as boundary condition determination information;
The boundary condition determination information and the residual orthogonal transform coefficient information after the inverse quantization are acquired, and obtained from the orthogonal transform coefficient information at a boundary between the rectangular area to be decoded and another adjacent rectangular area. The approximate orthogonal transform coefficient information obtained by orthogonal transform with respect to the estimation signal in the rectangular region to be decoded, which satisfies the Poisson equation with the gradient of the moving image signal as a boundary condition, is subjected to the inverse quantization. After generating using the residual orthogonal transform coefficient information, the decoded orthogonal transform coefficient information is obtained by combining the residual orthogonal transform coefficient information after the inverse quantization and the generated approximate orthogonal transform coefficient information. Generating inverse multiple harmonic local orthogonal transform means;
An inverse orthogonal transform unit that generates a decoded image signal for one screen by performing an inverse orthogonal transform on the decoded orthogonal transform coefficient information acquired from the inverse multiple harmonic local orthogonal transform unit, and the predetermined rectangular region A video decoding apparatus characterized in that decoding is performed in units of processing. The residual orthogonal transform coefficient information after dequantization obtained by the dequantization means is residual DCT coefficient information, and the boundary condition determination means determines that the residual when the boundary condition correction information is not valid. Based on the DCT coefficient information, the image signal discontinuity between the rectangular area to be decoded and the adjacent rectangular area is analyzed, and the boundary condition is corrected based on at least one predetermined determination condition. Means for generating boundary condition correction information for specifying whether or not boundary condition correction is necessary, and using the obtained boundary condition correction information as boundary condition determination information The inverse multiple harmonic local orthogonal transform means uses a DCT base as an orthogonal transform base, and performs orthogonal transform on the estimated signal in the rectangular region to be decoded so as to satisfy the Poisson equation. The generated approximate DCT coefficient information is generated using the residual DCT coefficient information after the inverse quantization, and then the residual DCT coefficient information obtained by the inverse quantization and the generated approximate DCT coefficient information are combined. 18. The moving picture decoding apparatus according to claim 16 , wherein the moving picture decoding apparatus is a means for generating decoded DCT coefficient information by performing. A coded bit stream generated by the moving picture coding apparatus according to claim 1 or 2 or the information processing apparatus operated by the moving picture coding program according to claim 7 or 8 is stored in a predetermined storage medium or a predetermined transmission path. A moving picture decoding program for causing a computer to execute a moving picture decoding to perform a decoding operation on the obtained encoded bitstream and output a decoded moving picture signal,
The computer,
Demultiplexing means for demultiplexing encoded information based on a predetermined syntax structure for the encoded bitstream;
Predemultiplexed information obtained from the demultiplexing means is subjected to predetermined entropy decoding, and post-quantization information, motion vector information, and parameter information necessary to construct a predetermined syntax structure Entropy decoding means to generate;
Inverse quantization means for generating residual DCT coefficient information as the residual orthogonal transform coefficient information after inverse quantization by performing inverse quantization on the quantized information acquired from the entropy decoding means; ,
A Poisson equation using a DCT basis as an orthogonal transformation basis and a gradient condition of a moving image signal obtained from the orthogonal transformation coefficient information at a boundary between the rectangular region to be decoded and another adjacent rectangular region as a boundary condition After generating the approximate DCT coefficient information orthogonally transformed with respect to the estimation signal in the rectangular region that is to be decoded, satisfying using the residual DCT coefficient information after the inverse quantization, An inverse multiple harmonic local orthogonal transform means for generating decoded DCT coefficient information by combining the residual DCT coefficient information obtained by inverse quantization and the generated approximate DCT coefficient information ;
An inverse orthogonal transform unit that generates a decoded differential image signal for one screen by performing an inverse orthogonal transform on the decoded DCT coefficient information acquired from the demultiplex harmonic local orthogonal transform unit;
A predicted moving image obtained by synthesizing a predicted image signal generated by predetermined motion compensation using the motion vector information acquired from the entropy decoding unit or a predicted image signal generated by intra prediction and the decoded difference image signal. A moving picture decoding program which functions as an image decoding means for generating a signal. An encoded bit stream generated by the moving image encoding apparatus according to claim 3 or 6 or the information processing apparatus operated by the moving image encoding program according to claim 9 or 12 is stored in a predetermined storage medium or a predetermined transmission path. A moving picture decoding program for causing a computer to execute a moving picture decoding to perform a decoding operation on the obtained encoded bitstream and output a decoded moving picture signal,
The computer,
Demultiplexing means for demultiplexing encoded information based on a predetermined syntax structure for the encoded bitstream;
Predemultiplexed information obtained from the demultiplexing means is subjected to predetermined entropy decoding to generate post-quantization information, motion vector information, and parameter information necessary for constructing a predetermined syntax structure. Entropy decoding means for extracting the boundary condition correction information encoded and incorporated, and
Inverse quantization means for generating the residual orthogonal transform coefficient information after inverse quantization by performing inverse quantization on the post-quantization information obtained from the entropy decoding means;
If the boundary condition correction information acquired from the entropy decoding means is not valid, the rectangle to be decoded is based on the post-quantization residual orthogonal transform coefficient information acquired from the inverse quantization means. Analyzing the discontinuity of the moving image signal between the region and the adjacent surrounding rectangular region, determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition, Generate boundary condition correction information for specifying whether or not condition correction is necessary, and use the generated boundary condition correction information or valid boundary condition correction information acquired from the entropy decoding means as a boundary condition Boundary condition determination means as determination information;
The boundary condition determination information and the residual orthogonal transform coefficient information after the inverse quantization are acquired, and the inclination of the orthogonal transform coefficient information at the boundary of the rectangular area to be decoded is adjacent to another rectangular area The approximate orthogonal transform coefficient information obtained by orthogonal transform for the estimated signal in the rectangular region to be decoded that satisfies the Poisson equation with the boundary condition as the residual orthogonality after the inverse quantization Demultiplexed harmonics that are generated using transform coefficient information and then generate decoded orthogonal transform coefficient information by combining the residual orthogonal transform coefficient information after the inverse quantization and the generated approximate orthogonal transform coefficient information A local orthogonal transform means;
An inverse orthogonal transform unit that generates a decoded difference image signal for one screen by performing an inverse orthogonal transform on the decoded orthogonal transform coefficient information acquired from the demultiplex harmonic local orthogonal transform unit;
A predicted moving image obtained by synthesizing a predicted image signal generated by predetermined motion compensation using the motion vector information acquired from the entropy decoding unit or a predicted image signal generated by intra prediction and the decoded difference image signal. A moving picture decoding program which functions as an image decoding means for generating a signal. The computer,
The residual orthogonal transform coefficient information after dequantization obtained by the dequantization means is residual DCT coefficient information, and the boundary condition determination means determines that the residual when the boundary condition correction information is not valid. Based on the DCT coefficient information, the image signal discontinuity between the rectangular area to be decoded and the adjacent rectangular area is analyzed, and the boundary condition is corrected based on at least one predetermined determination condition. Means for generating boundary condition correction information for specifying whether or not boundary condition correction is necessary, and using the obtained boundary condition correction information as boundary condition determination information Function as
The inverse multiple harmonic local orthogonal transform means uses DCT bases as orthogonal transform bases, and approximate DCT coefficient information obtained by performing orthogonal transform on the estimated signal in the rectangular region that is a decoding target that satisfies the Poisson equation. Is generated by using the residual DCT coefficient information after the inverse quantization, and is then decoded by combining the residual DCT coefficient information obtained by the inverse quantization and the generated approximate DCT coefficient information. 21. The moving picture decoding program according to claim 20 , wherein the moving picture decoding program functions as means for generating DCT coefficient information. Obtaining a coded bitstream generated by the moving picture coding apparatus according to claim 4 or the information processing apparatus operated by the moving picture coding program according to claim 10 from a predetermined storage medium or a predetermined transmission path; A moving picture decoding program that performs a decoding operation on the acquired encoded bitstream and outputs a decoded moving picture signal by a computer,
The computer,
Demultiplexing means for demultiplexing encoded information based on a predetermined syntax structure for the encoded bitstream;
The demultiplexed information obtained from the demultiplexing means is subjected to predetermined entropy decoding to generate post-quantization information and parameter information necessary for constructing a predetermined syntax structure, Entropy decoding means for extracting the boundary condition correction information incorporated into
Inverse quantization means for generating the residual orthogonal transform coefficient information after inverse quantization by performing inverse quantization on the post-quantization information obtained from the entropy decoding means;
If the boundary condition correction information acquired from the entropy decoding means is not valid, the rectangle to be decoded is based on the post-quantization residual orthogonal transform coefficient information acquired from the inverse quantization means. Analyzing the discontinuity of the moving image signal between the region and the adjacent surrounding rectangular region, determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition, Generate boundary condition correction information for specifying whether or not condition correction is necessary, and use the generated boundary condition correction information or valid boundary condition correction information acquired from the entropy decoding means as a boundary condition Boundary condition determination means as determination information;
The boundary condition determination information and the residual orthogonal transform coefficient information after the inverse quantization are acquired, and obtained from the orthogonal transform coefficient information at a boundary between the rectangular area to be decoded and another adjacent rectangular area. The approximate orthogonal transform coefficient information obtained by orthogonal transform with respect to the estimation signal in the rectangular region to be decoded, which satisfies the Poisson equation with the gradient of the moving image signal as a boundary condition, is subjected to the inverse quantization. After generating using the residual orthogonal transform coefficient information, the decoded orthogonal transform coefficient information is obtained by combining the residual orthogonal transform coefficient information after the inverse quantization and the generated approximate orthogonal transform coefficient information. Generating inverse multiple harmonic local orthogonal transform means;
And functioning as inverse orthogonal transform means for generating a decoded image signal for one screen by performing inverse orthogonal transform on the decoded orthogonal transform coefficient information acquired from the demultiplex harmonic local orthogonal transform means. A moving picture decoding program. An encoded bit stream generated by the moving image encoding device according to claim 5 or the information processing device operated by the moving image encoding program according to claim 11 is acquired from a predetermined storage medium or a predetermined transmission path, A moving picture decoding program that performs a decoding operation on the acquired encoded bitstream and outputs a decoded moving picture signal by a computer,
The computer,
Demultiplexing means for demultiplexing encoded information based on a predetermined syntax structure for the encoded bitstream;
The demultiplexed information obtained from the demultiplexing means is subjected to predetermined entropy decoding to generate post-quantization information and parameter information necessary for constructing a predetermined syntax structure, Entropy decoding means for extracting the boundary condition correction information incorporated into
By performing inverse quantization on the post-quantization information acquired from the entropy decoding means using a predetermined rectangular area as a processing unit, the residual orthogonal transform coefficient information after the inverse quantization is generated, and Dequantization means for outputting residual orthogonal transform coefficient information after dequantization in units of processing of the rectangular region;
When the boundary condition correction information acquired from the entropy decoding means is not valid, the boundary condition correction information is a decoding target based on the residual orthogonal transform coefficient information after the inverse quantization acquired from the inverse quantization means. Analyzing the discontinuity of the moving image signal between the rectangular area and the adjacent rectangular area, and determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition , Generating boundary condition correction information for specifying whether or not boundary condition correction is necessary, and generating the boundary condition correction information generated or the effective boundary condition correction information acquired from the entropy decoding means Boundary condition determination means as boundary condition determination information;
The boundary condition determination information and the residual orthogonal transform coefficient information after the inverse quantization are acquired, and obtained from the orthogonal transform coefficient information at a boundary between the rectangular area to be decoded and another adjacent rectangular area. The approximate orthogonal transform coefficient information obtained by orthogonal transform with respect to the estimation signal in the rectangular region to be decoded, which satisfies the Poisson equation with the gradient of the moving image signal as a boundary condition, is subjected to the inverse quantization. After generating using the residual orthogonal transform coefficient information, the decoded orthogonal transform coefficient information is obtained by combining the residual orthogonal transform coefficient information after the inverse quantization and the generated approximate orthogonal transform coefficient information. Generating inverse multiple harmonic local orthogonal transform means;
And functioning as inverse orthogonal transform means for generating a decoded image signal for one screen by performing inverse orthogonal transform on the decoded orthogonal transform coefficient information acquired from the demultiplex harmonic local orthogonal transform means. A moving picture decoding program. The computer,
The inverse quantization means functions as means for generating residual DCT coefficient information as residual orthogonal transform coefficient information after the inverse quantization,
When the boundary condition correction information is not valid, the boundary condition determination unit is configured to output an image signal between a rectangular area to be decoded based on the residual DCT coefficient information and an adjacent rectangular area. A boundary for analyzing the discontinuity, determining whether the boundary condition needs to be corrected based on at least one predetermined determination condition, and specifying whether the boundary condition needs to be corrected Condition correction information is generated, and the boundary condition correction information obtained is made to function as a boundary condition determination information.
The inverse multiple harmonic local orthogonal transform means uses DCT bases as orthogonal transform bases, and approximate DCT coefficient information obtained by performing orthogonal transform on the estimated signal in the rectangular region that is a decoding target that satisfies the Poisson equation. Is generated by using the residual DCT coefficient information after the inverse quantization, and is then decoded by combining the residual DCT coefficient information obtained by the inverse quantization and the generated approximate DCT coefficient information. The moving picture decoding program according to claim 22 or 23 , wherein the moving picture decoding program is caused to function as means for generating DCT coefficient information.

Images