JP2005184525A - Image processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor, having encoding and decoding functions of dynamic images with a reduced image compression data capacity, enabling compression/expansion processing for obtaining post-expansion image quality equivalent to an MPEG level or higher. <P>SOLUTION: In the image processor for encoding the dynamic image using I, P and B pictures, when an input original image is an I picture, the image processor includes an alternating-current predictive encoding section for dividing the original image into pixel blocks each having 4×4 pixels unit, obtaining a DC value which is a mean value of the pixel values, and performing alternating-current component prediction between neighboring pixel blocks; and when the input original image is a B or P picture, a motion estimation section for dividing the original image into macro blocks each having 8×8 pixels unit, extracting from other pictures a reference block having the strongest correlation, and deciding whether the encoded data of the motion compensation prediction of the macro block lies within a data capacity range, using the above extracted reference block. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to an image processing apparatus that performs image processing for creating image data (data structure for efficiently performing image compression) at a previous stage for compressing and expanding a moving image.

As a method for compressing moving images and sounds, there is a moving image and sound compression standard H.262 (commonly called “MPEG2” method) defined in Non-Patent Document 1.
In the moving picture compression by MPEG2, the image is divided into 16 × 16 pixel rectangular block groups called macroblocks, and is a region having a high correlation with the macroblock to be compressed from the temporally preceding and following images. (Hereinafter referred to as a reference region), the spatial distance and orientation (motion vector) from the reference region and the difference information between the reference region and the region to be compressed are calculated, and these information are converted into DCT (discrete). Cosine transform) and variable length coding to compress the bit stream.

The method of compressing only the motion vector and the difference information as described above can achieve much more efficient compression than the method of compressing the original image itself.
In other words, if the same degree of compression distortion is allowed, the information can be compressed to a smaller amount of information.
However, an image compressed with the difference information cannot be expanded without an image to be referenced.
For this reason, in order to cope with an application that extends from the middle of the bitstream (hereinafter referred to as “random access”), it is necessary to periodically provide images that do not refer to other images.

This image is referred to as an “I image” (Intra Picture), and the next B and P images are compressed using this I image as a reference image.
Among the images to be compressed using the reference image, a “P image (Predictive-coded Picture)” in which only a temporally previous image is used as a reference image, and a “B image” in which temporally preceding and succeeding images are used as reference images. There is “Image (Bidirectionally predictive-coded Picture)”. Similarly to the I image, the P image can be a reference image of another image, that is, a “B image”.
"ITU-T White Book, Audio Visual / Multimedia (H Series) Recommendation Collection" (Japan ITU Association, issued on February 18, 1995) P375-P595

However, in MPEG2 described above, image compression is performed independently in units of 16 × 16 pixel rectangular blocks.
That is, since the image compression is performed independently for each rectangular block in the case of an I image, there is a limit to improving the compression efficiency.
In addition, the B and P images also use difference data from the reference image and perform motion compensation to reduce the data amount. However, the rectangular block in which the reference image cannot be used is a rectangular block unit like the rectangular block of the I image. Therefore, there is a limit to improving the compression efficiency.

  An object of the present invention is made under such a background, and is to perform compression / decompression processing for reducing the capacity of compressed image data and obtaining image quality after decompression equivalent to or higher than that of MPEG. An object of the present invention is to provide an image processing apparatus having a function of encoding and decoding a possible moving image.

  The image processing apparatus according to the present invention is an image processing apparatus that encodes a moving image using an I picture, a P picture, and a B picture. When an input original image is an I picture, the original image is converted to 4 × 4. A 2 × 2 pixel sub-pixel block obtained by dividing the pixel block into pixel blocks, obtaining a DC value that is an average value of pixel values, performing AC component prediction between adjacent pixel blocks, and dividing each pixel block into four Each time, a first DC pixel value is obtained, and a difference between the first DC pixel value and a sub-pixel block of the original image at the corresponding position is obtained as a first difference value, and the first difference value is obtained. A first AC prediction encoding unit that encodes the first Hadamard coefficient and outputs the DC value and the first Hadamard coefficient as encoded data; and the pixel block is converted into a sub-pixel block in units of 2 × 2 pixels. Split Then, a DC value that is an average value of pixel values is obtained, AC component prediction is performed between adjacent sub-pixel blocks, a second DC pixel value is obtained for each pixel in each sub-pixel block, and the second DC The difference between the pixel value and the pixel in the sub-pixel block of the original image at the corresponding position is obtained as a second difference value, the second difference value is encoded as a second Hadamard coefficient, and the second Hadamard coefficient is When the input AC is a B or P picture, and the second AC predictive encoding unit that is output as encoded data, the original image is divided into macro block units of 8 × 8 pixels, A reference block having the highest correlation is extracted from another picture, and it is determined whether the encoded data for motion compensation of the macroblock using this reference block is within a predetermined data capacity range. A motion estimator, an inter coding unit that performs inter coding based on motion compensation for the macroblock when the coded data for motion compensation of the macroblock is within the predetermined range, and motion of the macroblock And an intra encoding unit that performs intra encoding on the macroblock when the compensation encoded data is not within the predetermined range.

  In the image processing apparatus of the present invention, the inter coding unit obtains a difference value between an 8 × 8 pixel reference image (reference block) extracted by motion prediction and a macroblock to be coded, and uses the difference value. The difference block is divided into 4 × 4 pixel blocks (residual blocks), the difference value is averaged for each pixel block to obtain a DC value, and the difference between the DC value and the difference value of the difference block is calculated. Are encoded as Hadamard coefficients, and the DC value and Hadamard coefficients are output as encoded data.

  In the image processing apparatus of the present invention, the intra encoding unit divides a macroblock to be encoded into pixel blocks of 4 × 4 pixel units, obtains a DC value that is an average value of pixel values, and sets adjacent pixel blocks The first DC pixel value is obtained for each 2 × 2 pixel sub-pixel block obtained by dividing each pixel block into four, and the first DC pixel value and the macro block are The difference from the sub-pixel block at the corresponding position is obtained as a first difference value, the first difference value is encoded as a first Hadamard coefficient, and the DC value and the first Hadamard coefficient are used as encoded data. A third AC predictive encoding unit for output and the pixel block are divided into sub-pixel blocks in units of 2 × 2 pixels, and a DC value that is an average value of the pixel values is obtained and placed between adjacent sub-pixel blocks. AC component prediction is performed to obtain a second DC pixel value for each pixel in each sub-pixel block, and a difference between the second DC pixel value and a pixel in the sub-pixel block of the original image at the corresponding position is calculated. And a fourth AC predictive encoding unit that obtains the second difference value, encodes the second difference value as a second Hadamard coefficient, and outputs the second Hadamard coefficient as encoded data. And

  In the image processing apparatus according to the present invention, the AC prediction encoding unit divides the macroblock to be encoded into pixel blocks of 4 × 4 pixel units, and further divides the macroblocks into 2 × 2 pixel sub-pixel blocks. When performing the conversion, if the adjacent target pixel block is an inter-coded pixel block, the AC component prediction is performed using the macroblock of the decoding result.

  The image processing apparatus according to the present invention is an image processing apparatus that performs decoding of a moving image using an I picture, a P picture, and a B picture. When the input encoded image data is an I picture, the input decoding is performed. For each 2 × 2 pixel sub-pixel block obtained by performing AC component prediction between the DC value of the target pixel block of interest and the DC value of the pixel block adjacent to the target pixel block, and dividing each pixel block into four, A first DC pixel value is obtained, an inverse Hadamard transform of the input first Hadamard coefficient is performed on the first DC pixel value, and the first DC pixel value corresponding to the first DC pixel value of the sub-pixel block is obtained. The difference from the sub-pixel block of the original image at the position is obtained as a first difference value, and this first difference value is added to the first DC pixel value decoded by the AC component prediction, and the addition result AC component prediction is performed between the first AC prediction decoding unit that outputs the decoded data, and the first DC pixel value and the first DC pixel value of the sub-pixel block adjacent to the first DC pixel value. For each pixel, a second DC pixel value is obtained, an inverse Hadamard transform of the input second Hadamard coefficient is performed on the second DC pixel value of this pixel, and a second DC of this pixel is obtained. The difference between the pixel value and the pixel of the original image at the corresponding position is obtained as a second difference value, and this second difference value is added to the second DC pixel value decoded by the AC component prediction, The second AC predictive decoding unit that outputs the addition result as decoded data, and when the encoded image data to be input is a B or P picture, a macro block of 8 × 8 pixels to be decoded is intra-coded. Or inter-coded When the decoding unit and the decoding target macroblock are inter-coded, the difference block based on the difference between each pixel of the reference block and the macroblock is divided into 4 × 4 pixel blocks. The residual DC value, which is the average value of the residuals obtained for each pixel block, and the first and second Hadamard coefficients obtained for each 4 × 4 pixel block, An inter decoding unit that decodes the macroblock by adding the difference value of the residual obtained by inverse Hadamard transform of the second Hadamard coefficient and adding the pixel value of the reference block to the addition result; and decoding When the target macroblock is inter-coded, the residual DC value, which is the average of the residuals of the pixels of the reference block and the macroblock, and the macroblock to be decoded are An intra-decoding unit that obtains a DC value of a pixel block of 4 × 4 pixels from an input macroblock and decodes the macroblock based on the DC value when it is a tra-coded one .

  In the image processing apparatus according to the present invention, the intra decoding unit divides a macroblock to be decoded into pixel blocks of 4 × 4 pixel units, and further divides the macroblocks into 2 × 2 pixel sub-pixel blocks. When performing decoding, if the adjacent target pixel block is an inter-coded pixel block, AC component prediction is performed using a macroblock of the decoding result.

  According to the image processing apparatus of the present invention, the first difference value between the pixel value of the original image and the first DC pixel value of the sub-pixel block is subjected to Hadamard transform, and a low-frequency component having a large amount of information as the difference value ( After extracting the first Hadamard coefficient corresponding to the pixel block, the first difference value is decoded and added to the first DC pixel value to obtain the second DC pixel value of each pixel of the sub-pixel block. Since the Hadamard coefficient (second Hadamard coefficient corresponding to the sub-pixel block) on the remaining high frequency side is obtained by the second Hadamard transform, the information amount of the remaining second Hadamard coefficient can be reduced. Compared with the one-stage process in which the ACP process and the Hadamard transform are performed once, the data amount of the second difference value can be reduced, and all Hadamard coefficients are obtained from the one-stage difference value. Compared to the case, to reduce the data amount of the Hadamard coefficients, the image quality of MPEG and more equal, it is possible to perform the compression and decoding of the image.

Also, according to the image processing apparatus of the present invention, not only Hadamard transform but also encoding processing by ACP method is added, so that block noise that tends to appear because Hadamard transform is performed in units of blocks can be reduced.
Furthermore, according to the image processing device of the present invention, at the time of encoding, inter decoding is performed, and the decoding result of inter is used for intra encoding processing, that is, to obtain a difference value for encoding. Therefore, the above-described image compression method for compressing the difference value with the image to be compared can be effectively applied to the moving image, and the compression efficiency can be greatly improved and transmitted as moving image information. The amount of encoded data to be reduced can be reduced, the transfer efficiency of image information can be improved, and an image can be reproduced with an image quality equivalent to MPEG by decoding the encoded data.

In general, the capacity of image data is very large, but it has a lot of redundancy (a highly correlated part).
In still images, adjacent pixels often take very similar values (high spatial correlation).
Further, similar images appear in succession in moving images, although there are some differences (the temporal correlation is high).
By effectively removing the spatial and temporal redundancy described above, it is possible to significantly compress image information.

When spatial correlation is used If an adjacent pixel block having a high correlation with the encoding target block is detected, the value of the pixel block to be encoded next from the adjacent pixel block that has already been encoded is set. Information compression can be performed by encoding the difference between the pixel value of the adjacent pixel block and the pixel value of the pixel block to be encoded.

When using temporal correlation In a moving image, continuous images often have a very high correlation, that is, time-dependent changes are often limited.
The simplest technique for reducing redundancy is to take the difference (B−A) between each pixel of the image (A) encoded immediately before and each pixel of the image (B) to be encoded. There is an interframe differential encoding that encodes.
This inter-frame difference encoding method is effective when the motion is small, but when the change between frames is large, the code amount may increase as compared to the case of encoding as it is.
When the amount of code increases from a predetermined set value, the block having the smallest error is searched for the block most similar to the pixel block in the image to be encoded from the already encoded image. There is motion compensation that encodes the extracted position and the difference between the extracted position and the pixel block to be encoded.
In the image processing apparatus of the present invention, a moving image is encoded by effectively using the above-described spatial correlation and temporal correlation (motion compensation method).

In addition, as described above, the motion compensation divides the encoding target image into 8 × 8 pixel macroblocks on the rectangle of the pixel block, and the macroblock with the highest correlation is referred to as the reference block for each encoding target block. The difference between the block to be encoded and the reference block is obtained, and this difference is encoded.
However, when the scene of the image changes or when a new object appears in the image, this motion compensation cannot be used. Therefore, encoding efficiency increases when the pixel block is encoded in the same manner as a still image.

For this reason, pixel blocks are classified into the following two types and subjected to encoding processing.
・ Inter block (inter-coded macro block)
A block in which there is a reference block having a high correlation in the reference image, and a difference from this reference block is encoded.
Intra block (intra-coded macro block)
A block that is encoded like a still image without a reference block having a high correlation in the reference image.

Motion compensation is an indispensable technique for high-compression coding of moving images. However, in order to decode a motion-compensated image frame, it is necessary that a reference image used for motion prediction has already been decoded.
As the number of reference images increases, handling of the buffer for storing the images and the delay of time until display becomes more complicated, and in order to eliminate this complexity, the present invention, like MPEG, has the following three types. Images are classified into picture types.

In an I-picture (Intra-coded Picture), all pixel blocks are encoded as intra blocks. For this reason, the I picture is not related to other images, can be decoded alone, and is used to restore the subsequent P and B pictures after decoding.
P (Predictive-coded Picture) is restored by forward motion compensation on the time axis using a past I picture or P picture as a reference picture. For this reason, the P picture depends only on (related to) the past I or P picture, and is used to restore the subsequent P and B pictures after decoding.
B (Bidirectional-coded picture) is restored by forward and backward motion compensation on the time axis by referring not only to past I and P pictures but also to future I and P pictures as reference pictures.

<< Encoding Unit >>
The image processing apparatus of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram illustrating a configuration example of an image processing apparatus that performs encoding according to an embodiment of the present invention.
This image processing apparatus has a picture rearrangement unit 1, a D-RAPIC encoding unit 2, a picture storage memory 3, a motion estimation unit 4, an inter encoding unit 5, an intra encoding unit 6, and a variable length encoding unit 7. ing.
The picture rearrangement unit 1 classifies an original image (frame image) input from an external device in time series in units of frames into three types of I, P, and B pictures, and classifies them into I, P, and B pictures. As shown in FIG. 2, each image is given an identifier indicating whether it is an I, P or B picture, and rearranged in the order of encoding according to this identifier (encoding process). Here, the above-described rearrangement of pictures is performed because it is necessary to refer to a P picture that exists in the future from itself in the time-series arrangement displayed when picture B is encoded. As an example of the classification rule of the original image into I, P, and B pictures, the cycle N of the I picture is set every 9 frames, and the cycle M of the P picture is set every 3 frames. Here, for example, the period M is normally set to “3”, and the period N needs to be a multiple of the period M. In the case of 30 fps (frames / sec), the period N is “15”, and in the case of 25 fps, The period N is “12”. Further, the classification rule for the original image into I, P, and B pictures is not limited to the above combination, but is set as appropriate.

  In the configuration of FIG. 1, the picture rearrangement unit 1 includes an input permutation [0 (I), 1 (B), 2 (B), 3 (P), 4 (B), 5 (B), 6 ( P), 7 (B), 8 (B), 9 (I), 10 (B), 11 (B), 12 (P)] are encoded permutations that are in the order of encoding according to the identification. [0 (I), 3 (P), 1 (B), 2 (B), 6 (P), 4 (B), 5 (B), 9 (I), 7 (B), 8 (B) , 12 (B), 10 (B), 11 (B)], the type of the picture is determined by the identifier, and it is determined that the original image is an I picture, When the image is output to the D-RAPIC encoding unit 2 and it is determined that the original image is a B or P picture, the original image converted to the B or P picture is output to the motion estimation unit 4.

The D-RAPIC encoding unit 2, which will be described in detail later, encodes an I picture by an operation based on an ACP (alternating current component prediction) method, and sends the encoded data to the variable length encoding unit 7. Output.
The motion estimation unit 4 sequentially decomposes the P and B pictures into macroblocks (8 × 8 pixel blocks) from the upper left end, and calculates a motion vector for each macroblock unit.
Here, the motion estimation unit 4 uses an I picture and a P picture as reference images for a P picture, and a reference region having a high spatial correlation with a macro block to be encoded in the P picture (hereinafter referred to as an encoding target block). Are searched from the I picture and the P picture, and the reference region having the highest spatial correlation is extracted from either the I picture or the P picture.

In addition, the motion estimation unit 4 uses I and P pictures as reference images for the B picture, and selects a pixel block to be encoded in the B picture (hereinafter referred to as an encoding target block) and a reference region having a high spatial correlation. Search from the I and P pictures, and extract the reference region with the highest spatial correlation than either the I or P picture.
As a search method, as in the general method, the minimum value of the prediction calculation difference signal is obtained by shifting horizontally and vertically within a predetermined pixel range.

  At this time, the motion estimator 4 calculates a motion vector that minimizes the prediction residual signal between the extracted reference region and the encoding target block (the sum of absolute values and the sum of squares is minimized). It is determined whether or not the obtained minimum predicted arithmetic difference signal is equal to or less than a predetermined residual threshold value (within a predetermined data capacity range). Based on the determination result, the encoding method shown below is determined. Macro block coding information (MCB coding information) indicating which coding method (inter coding or intra coding) is used for each coding target block is selected. ) Is attached.

At this time, if the prediction residual signal is equal to or less than the residual threshold, the motion estimation unit 4 can perform encoding with higher efficiency when encoded as a motion compensation block using a motion vector, that is, as a still image. It is determined that the amount of encoded data is smaller than that of the encoded data, and this encoding target block is set as an inter block, which is output to the inter encoding unit 5 and macro block encoding information indicating that it is an inter block. And output to the variable length encoding unit 7.
At this time, the motion estimation unit 4 outputs the motion vector to the variable length coding unit 7.

On the other hand, when the prediction residual signal exceeds the residual threshold, the motion estimation unit 4 can perform encoding with higher efficiency when encoded as a still image, i.e., encoded than as encoded as a still image. The encoding target block is determined to be an intra block, and is output to the intra encoding unit 6. Macro block encoding information indicating the intra block is also transmitted to the variable length encoding unit 7. Output.
Here, the motion estimation unit 4 sets the macroblock coding information to a value that distinguishes the following three cases.
-Intra-coded (intra-block)-When inter-coded with reference to past pictures-When inter-coded with reference to future pictures

  Hereinafter, the configuration and operation examples of the D-RAPIC encoding unit 2, the inter encoding unit 5, and the intra encoding unit 6 according to the embodiment of the present invention will be sequentially described with reference to the drawings.

<D-RAPIC encoding unit 2>
FIG. 3 is a block diagram showing a configuration example of the D-RAPIC encoding unit 2 according to an embodiment of the present invention.
First, the outline of the processing in the D-RAPIC encoding unit 2 will be described. The image processing apparatus of the present invention divides an original image into pixel blocks composed of a plurality of pixels (n × n pixels), for example, 4 × 4. Are divided into pixel blocks each consisting of a pixel.
Then, the average value of the pixel values of each pixel is calculated for each pixel block, and in this pixel block, the DC value of each DC image (that is, the pixel value in the pixel block) is calculated as the averaged pixel value of each pixel. An average value of pixel values) is obtained.
Then, the image processing apparatus according to the present invention uses the first value for each sub-pixel block (block of 2 × 2 pixels) from the calculation based on the first-stage ACP (AC component prediction) method based on the DC value. A DC pixel value is obtained, and the first difference value is obtained by subtracting the first DC pixel value from the average value of the pixel values of the pixels of the original image in each sub-pixel block, and the first difference value is obtained. Then, Hadamard transform is performed, a part of the Hadamard coefficient is obtained, quantized, and output.

Next, the D-RAPIC encoding unit 2 performs inverse quantization on the first Hadamard coefficient, performs inverse Hadamard transformation, and reproduces the first difference value (the information reproduced by the quantization is deteriorated). )
Then, the D-RAPIC encoding unit 2 adds the first difference value to the corresponding pixel value of the sub-pixel lock (that is, the first DC pixel value), and a new first value of the sub-DC image is added. Output as a DC pixel value.
Next, the D-RAPIC encoding unit 2 performs an operation based on the ACP method in the second stage based on the new first DC pixel value of the DC image, and the second DC pixel for each pixel. Find the value.

Then, the D-RAPIC encoding unit 2 obtains a second difference value between the pixel value of each pixel of the original image and the second DC pixel value (the second DC pixel value is obtained from the pixel value of the pixel). Subtraction), the second Hadamard transform is performed, and the remaining Hadamard coefficients excluding the first Hadamard coefficient (AC code) are obtained as second Hadamard coefficients (AC code).
As described above, the D-RAPIC encoding unit 2 similarly obtains the Hadamard coefficient by dividing the difference value between the pixel value obtained by the ACP method in the above two steps and the pixel value of the original image into two corresponding steps. An operation is performed.

In this way, the D-RAPIC encoding unit 2 generates a DC value at the time when the DC image processing is performed on the 16 pixels, and the low-frequency component is 3 in the first-stage Hadamard transform. The first Hadamard coefficients are obtained, and in the second-stage Hadamard transform, three high-frequency components are obtained for every four subpixel blocks, and a total of twelve second Hadamard coefficients are obtained.
Then, the D-RAPIC encoding unit 2 performs entropy encoding (for example, Huffman) after quantizing the obtained 15 Hadamard coefficients and DC values, and outputs the data after compression.

  Thereby, in the first Hadamard transform, the D-RAPIC encoding unit 2 calculates the first difference value between the average value of the pixel values of the original image in the sub-pixel block range and the first DC pixel value. After converting, a low frequency component with a large amount of information is extracted as a difference value, encoded as a first Hadamard coefficient by Hadamard transform, and further compressed and encoded by quantization, and then the first difference value is encoded. Decoding by inverse quantization and further decoding by inverse Hadamard transform to obtain a new first DC pixel value in addition to the first DC pixel value, and from this new first DC pixel value to the second DC After obtaining the pixel value, it is encoded as a second Hadamard coefficient by the second Hadamard transform, and the remaining high-frequency Hadamard coefficient is obtained.

  On the other hand, since the D-RAPIC encoding unit 2 encodes image data by Hadamard transform using a rectangular function as a basis function, a sine function is assumed when an image in which the pixel values of adjacent pixels change smoothly is assumed. Compared with DCT having a basis function of, information regarding a smooth change cannot be efficiently encoded.

  Further, the D-RAPIC encoding unit 2 needs to be quantized from the viewpoint of compression thereafter, and when performing inverse Hadamard transform after inverse quantization at the time of image decoding, the DCT can be restored by approximating smoothly. On the other hand, because of the square wave in the Hadamard transform, the edge is sharp and the pixel value difference between the pixels is easy to open.

  However, the D-RAPIC encoding unit 2 extracts the first Hadamard coefficient related to the low-frequency component by the first-stage Hadamard transform, and then adds the coefficient value to the first DC pixel value to create a new first DC pixel. This value is reflected in the encoding, and the first DC pixel value is compensated for the difference from the surrounding block by ACP, the difference from the pixel value of the original image is obtained, and the second difference value is the second value. By performing the step Hadamard transform, the image is restored with an image quality equivalent to or better than that of the DCT, and the difference value from the pixel value of the original image is taken. Therefore, the second step processing (the second step ACP and the second step) Encoding with high accuracy is possible in stage Hadamard transform).

Further, the D-RAPIC encoding unit 2 reflects the result of the first stage process (first stage ACP and first stage Hadamard transform), and the second stage process (second stage ACP and second stage Hadamard transform). ) Can be suppressed, and the mosquito noise that appears when the data is highly compressed and restored by DCT or the like can be greatly reduced by the ACP process.
Furthermore, in DCT, a computation with a large load of real number accuracy is required, but the D-RAPIC encoding unit 2 is a combined method of ACP and Hadamard transform, and is required for encoding by adding / subtracting bit shift to an integer value. Therefore, compression / decompression can be performed at a higher speed than DCT.

The D-RAPIC encoding unit 2 also includes a DC value for performing the ACP process, first and second DC pixel values after the ACP process is performed on the DC value, and a sub-pixel block in the original image. Since the Hadamard coefficient of the information obtained by Hadamard transforming the difference value between each of the DC value and the pixel value is encoded and output, the amount of information necessary for the image reproduction processing is reduced as described above. Since the difference value from the image is output as a Hadamard coefficient, the accuracy of the encoded data (image quality at the time of decoding, that is, the accuracy of reproduction) can be improved, and it is less than the DCT processing used in MPEG or the like. It is possible to give high reproducibility to the original image with the amount of information.
The first DC value and the second DC value can be easily obtained by averaging each pixel as described above if there is a DC value (DC image). Is not output.

Next, the configuration and operation of the D-RAPIC encoding unit 2 will be described with reference to FIG.
The D-RAPIC encoding unit 2 decomposes the original image into pixel blocks (4 × 4 pixels) in order from the upper left end, generates a DC image from the pixel block, and generates a DC image based on the DC image. A first-stage AC prediction component that performs one-stage ACP processing, decomposes the pixel block (divides into four), creates sub-pixel blocks (2 × 2 pixels), and obtains a first DC pixel value for each sub-pixel block A first difference value between the prediction unit 104, the first DC pixel value, and the average value of the pixel values of the original image within the sub-pixel block range is obtained, and a first step is performed on the first difference value. The first Hadamard transform (AC code) is obtained and output, and the first DC pixel value and the first difference value are added to generate a new first DC pixel value. The first-stage Hadamard encoding unit 105 that outputs Based on the new first DC pixel value, the second stage AC component prediction unit 106 that performs the second stage ACP processing and calculates the second DC pixel value corresponding to each pixel of the original image; A second difference value between the pixel value of each pixel and the second DC pixel value is obtained, a second Hadamard transform is performed on the second difference value, and a second Hadamard coefficient (AC code) is obtained. The second stage Hadamard encoding unit 107 is obtained.

Next, an operation example of the D-RAPIC encoding unit 2 according to an embodiment of the present invention will be described with reference to FIG.
The D-RAPIC encoding unit 2 of the present invention divides the original image (2a in FIG. 3) into pixel blocks composed of, for example, 4 × 4 pixels in order from the upper left end, and the pixels in the pixel block are divided. An average value of pixel values is obtained as a DC value, and is output as a DC code after quantization.

A pixel block whose pixel value is indicated by this DC value is defined as a DC image.
That is, as shown in FIG. 4A, 4 × 4 pixels are used as pixel blocks, and as shown in FIG. 4B, each pixel block becomes a DC image by obtaining a DC value.
Here, the pixel block S includes 16 pixels d (0,0) to d (3,3) (these pixels respectively correspond to the pixel values d00 to d33 in FIG. 4D). A DC value that is an average value of pixel values of each pixel is used as the pixel value.

The first stage AC component prediction unit 104 generates a sub-pixel block from the DC image based on the AC component prediction method.
That is, in FIG. 5A, the first-stage AC component prediction unit 104 starts from the pixel block S and starts from the upper left end, and the sub-pixel blocks S1, S2, S3, and S4 (2nd pixel blocks S1, S2, S3, and S4) When the DC value (pixel value as a sub-pixel block, that is, the first DC pixel value) of the one-step ACP-applied image 2c) is predicted (generated) by the calculation described later, The first DC pixel value, together with the DC value of the pixel block S, the DC values of the pixel blocks (DC images) U, L, R, and B that are adjacent to the pixel block S in the vertical and horizontal directions (this DC value is the pixel Using the same code as the block).
That is, in the AC component prediction method, when S is a pixel block including sub-pixel blocks S1 to S4, the first DC pixel value of each of the sub-pixel blocks S1, S2, S3, and S4 is a sub-pixel block that performs prediction. Is obtained by the following equation using an upper pixel block U, a lower pixel block B, a left pixel block L, and a right pixel block R adjacent to the pixel block S having
S1 = S + (U + L-B-R) / 8
S2 = S + (U + R-B-L) / 8
S3 = S + (B + L-U-R) / 8
S4 = S + (B + R-UL) / 8

The first-stage Hadamard encoding unit 105 corresponds to each of the sub-pixel blocks S1, S2, S3, and S4, so that the pixels d (0,0), d (0,1), d (1, 0), d (1,1), pixels d (0,2), d (0,3), d (1,2), d (1,3) and pixels d (2,0), d (2,1), d (3,0), d (3,1) and pixels d (2,2), d (2,3), d (3,2), d (3,3) The average value of the pixel values of each of the four pixels is calculated.
The first-stage Hadamard encoding unit 105 subtracts the average pixel value of the corresponding pixel from the first DC pixel values of the sub-pixel blocks S1, S2, S3, and S4, The first difference value at is calculated, and a first-stage Hadamard transform is performed on the first difference value.

この（１）式において、Ｈはアダマール変換に用いるアダマール変換係数行列であり、以下に示す（２）式の構造となっている。

また、Ｈ^Ｔは、Ｈに対する転置行列であり、（２）式の構造と同様である。 The first-stage Hadamard encoding unit 105 regards the configuration of the sub-pixel block as 2 × 2 pixels as shown in the following equation (1), and the first-order Hadamard encoding unit corresponding to the 2 × 2 sub-pixel block: Perform conversion.

In this equation (1), H is a Hadamard transform coefficient matrix used for Hadamard transform, and has the structure of the following equation (2).

Also, H ^T is a transposed matrix for H, is the same as the structure of the formula (2).

ここで、ｓ(x,y)は、サブ画素ブロックに対応した各々の第１の差分値の値であり、変換係数ｈu,v(x,y)もサブ画素ブロックに対応して設けられている。 Α indicates a matrix of calculated Hadamard coefficients, and f corresponds to a difference value of each pixel described below corresponding to 2 × 2 pixels (that is, a first difference value of each sub-pixel block). .
Each element αu, v of the matrix of α is obtained by the following equation (3).

Here, s (x, y) is the value of each first difference value corresponding to the sub-pixel block, and the conversion coefficient hu, v (x, y) is also provided corresponding to the sub-pixel block. Yes.

  The transform coefficient hu, v (x, y) in the Hadamard transform in the 2 × 2 subpixel block corresponds to the transform coefficient shown in FIG. 6A (white is “−1”, black is "+1"). Here, the positional relationship (in the conversion coefficient shown in FIG. 6A) of the conversion coefficients hu, v (x, y) corresponding to each of the sub-pixel blocks S1, S2, S3, and S4 is shown in FIG. This corresponds to the positional relationship indicated by x, y of s (x, y) shown (only the positional relationship of the conversion coefficients hu, v (x, y) is shown here).

この第１段階のＡＣＰ後においては、画素ｄ(0,0)〜ｄ(3,3)の画素値は、帰属するサブ画素ブロック，すなわち対応するサブ画素ブロックＳ1，Ｓ2，Ｓ3，Ｓ4の第１のＤＣ画素値に対応している。 Moreover, the matrix of f has the structure of the following (4) Formula.

After the first-stage ACP, the pixel values of the pixels d (0,0) to d (3,3) are assigned to the subpixel blocks to which they belong, that is, the corresponding subpixel blocks S1, S2, S3, S4. This corresponds to a DC pixel value of 1.

  At this time, the first-stage Hadamard encoding unit 105 obtains the Hadamard coefficients α01, α02, α03 excluding the DC value and the second Hadamard coefficient generated by the second-stage Hadamard transform, so that the sub-pixel blocks S1, S2, The first difference value (s (0,0), s (0,1), s (1,0), s (1,1)) corresponding to each of S3 and S4 is used as the pixel d (0 , 0), d (0,1), d (1,0), d (1,1), the subpixel block S1 DC value is subtracted from the average pixel value of the pixel d (0,2) of the original image. , D (0,3), d (1,2), d (1,3), the subpixel block S2 DC value is subtracted from the average pixel value of the pixel d (2,0), d ( 2,1), d (3,0), d (3,1) subtracts the DC value of the sub-pixel block S3 from the average pixel value to obtain pixels d (2,2), d (2,3) of the original image. ), D (3,2) and d (3,3) are obtained by subtracting the DC value of the sub-pixel block S4 from the average pixel value.

  That is, in the 2 × 2 subpixel blocks of the subpixel blocks S1, S2, S3, and S4, the subpixel block is regarded as one pixel, and 2 × 2 Hadamard transform is performed on the first difference value of the subpixel block. .

The Hadamard coefficients α01, α10, and α11 obtained by Expression (3) are quantized by the first-stage Hadamard encoding unit 105 and output as first Hadamard coefficients.
Further, the first-stage Hadamard encoding unit 105 performs inverse quantization on the quantized first Hadamard coefficient, further performs inverse Hadamard transform, reproduces the first difference value, and reproduces the reproduced first difference value. Are added to the first DC pixel value corresponding to each pixel, and output as a new first DC pixel value in the image after application of the first stage Hdm (Hadamard transform) (2d in FIG. 3).
The information of the first difference value to be reproduced is deteriorated when quantized.

The second-stage AC component prediction unit 106 inputs the first DC pixel value of each of the sub-pixel blocks S1, S2, S3, and S4 of the input first-stage Hdm applied image, and the vertical and horizontal directions of each sub-pixel block. Based on the first DC pixel value of the adjacent sub-pixel block, the calculation of the AC component prediction is performed, and the second DC pixel value dd00 of each pixel d (0,0),..., D (3,3). ˜dd33 is obtained and an image after applying the second stage ACP (2e in FIG. 3) is output.
Here, the second-stage AC component prediction unit 106, for example, the second DC pixel value dd00 of each of the pixels d (0,0) to d (1,1) in the sub-pixel block S1 shown in FIG. , Dd01, dd10, dd11 are calculated by the following formula.
That is, in the AC component prediction method, when S1 is the first DC pixel value relating to the pixels d (0,0) to d (1,1), the second DC pixel values dd00, dd01, and dd10 of the respective pixels. , dd11 is obtained by the following equation using the first pixel values of the upper subpixel block U3, the lower subpixel block S3, the left subpixel block L2, and the right subpixel block S2.
dd00 = S1 + (U3 + L2-S3-S2) / 8
dd01 = S1 + (U3 + S2-S3-L2) / 8
dd10 = S1 + (S3 + L2-U3-S2) / 8
dd11 = S1 + (S3 + S2-U3-L2) / 8
Similarly, the second-stage AC component prediction unit 106 calculates the second DC pixel value of each pixel in the remaining remaining sub-pixel blocks S2 to S4, and calculates the second value of each pixel from the sub-pixel blocks S1 to S4. Are output as an image after application of the second stage ACP.

  The second-stage Hadamard encoding unit 107 subtracts the pixel value of the corresponding pixel from the second DC pixel value of each pixel of the sub-pixel blocks S1, S2, S3, and S4, and performs the first corresponding to each pixel. The second difference value is calculated, and a second-stage Hadamard transform is performed on the second difference value.

この（５）式において、Ｋnはアダマール変換に用いるアダマール変換係数行列であり、４×４画素の画素ブロック内の各画素のアダマール変換を行う場合でも、本発明においては２×２画素のサブ画素ブロック単位で各画素のアダマール変換を行うため、以下に示す（６）式の構造となっている。

The second-stage Hadamard encoding unit 107 performs Hadamard transform on the second difference value for each pixel in each sub-pixel block as 4 × 4 pixels in the following equation (5).

In Equation (5), Kn is a Hadamard transform coefficient matrix used for Hadamard transform, and even when Hadamard transform of each pixel in a 4 × 4 pixel block is performed, in the present invention, 2 × 2 pixel sub-pixels are used. Since the Hadamard transform of each pixel is performed in block units, the structure of the following formula (6) is adopted.

また、以下の説明において、画素ｄ(0,0)〜ｄ(3,3)の画素値を、対応する部分の符号に対応させて、図４（ｄ）に示すように画素値ｄ00〜ｄ33（対応関係は図５(ｂ)）で表している。
すなわち、画素ｄ(0,0)〜ｄ(3,3)の画素値は、以下に示すように、対応するサブ画素ブロックＳ1，Ｓ2，Ｓ3，Ｓ4における各画素値の第２のＤＣ画素値に対応している。 Βn represents a matrix of the obtained Hadamard coefficients, and gn represents a difference value (that is, a difference value of each pixel for each sub pixel block corresponding to 2 × 2 pixels, as shown in the following equation (7)). , Corresponding to a matrix of second difference values for each pixel in each sub-pixel block).

Further, in the following description, the pixel values of the pixels d (0,0) to d (3,3) are made to correspond to the codes of the corresponding portions, and the pixel values d00 to d33 are shown in FIG. (The correspondence is shown in FIG. 5B).
That is, the pixel values of the pixels d (0,0) to d (3,3) are the second DC pixel values of the respective pixel values in the corresponding sub-pixel blocks S1, S2, S3, and S4 as shown below. It corresponds to.

At this time, the second-stage Hadamard encoding unit 107 uses the pixels d (0,0), d (0,1), d (1,0), d in each of the sub-pixel blocks S1, S2, S3, and S4. (1,1), pixels d (0,2), d (0,3), d (1,2), d (1,3), pixels d (2,0), d (2,1), Each second DC pixel value of d (3,0), d (3,1), pixel d (2,2), d (2,3), d (3,2), d (3,3) And pixels d (0,0), d (0,1), d (1,0), d (1,1), pixels d (0,2), d (0,3), d of the original image (1,2), d (1,3), pixel d (2,0), d (2,1), d (3,0), d (3,1), pixel d (2,2), The difference between each pixel value of d (2,3), d (3,2), and d (3,3) is calculated (from the pixel value (d00 to d33) of the original image, the second value of each corresponding pixel is calculated. For example, the second DC pixel value dd00 is subtracted from the pixel value d00 of the original image at the pixel d (0,0), and this is d (0,0) -d ( 3), and the calculation result is set as the second difference value as shown in FIG. As shown, for each sub-pixel block (S1 to S4), en (0,0), en (0,1), en (1,0), en (1,1), n∈ [0,3] is obtained, and the second-stage Hadamard transform is performed by the following equation (8) using the second difference value.
Here, the subscript “n” of kn, βn, gn, en (x, y) corresponds to the position of each sub-pixel block. For example, referring to FIG. 4E, “n = 0” corresponds to the sub-pixel block S1, “n = 1” corresponds to the sub-pixel block S2, and “n = 2” corresponds to the sub-pixel block. Corresponding to S3, “n = 3” corresponds to the sub-pixel block S4.

ここで、２×２画素におけるアダマール変換係数ｋn(u,v)(x,y)は、図６（ａ）に示す各画素に対応した構成となっている（白が「−１」であり、黒が「＋１」である）。ここで、図６（ａ）におけるアダマール変換係数ｋn(u,v)(x,y)の位置関係は、図６（ｂ）に示している。 Then, each element βn (u, v) of the matrix of βn is obtained by the following equation (8).

Here, the Hadamard transform coefficient kn (u, v) (x, y) in 2 × 2 pixels has a configuration corresponding to each pixel shown in FIG. 6A (white is “−1”). Black is “+1”). Here, the positional relationship of the Hadamard transform coefficients kn (u, v) (x, y) in FIG. 6 (a) is shown in FIG. 6 (b).

Further, in the Hadamard transform coefficient of the present invention shown in FIG. 6A, the transform coefficient portion corresponding to the first-stage Hadamard transform described above is the same as the conventional configuration, but the second-stage Hadamard transform is used. The corresponding part is configured to correspond to each sub-pixel block (S1, S2, S3, S4).
That is, in a pixel block composed of 4 × 4 pixels, four groups A in FIG. 6A (in A1 to A4, A1 is h0,0 (x, y) and A2 is h0,1 (x, y). , A3 is h1,0 (x, y), and A4 is h1,1 (x, y)).
Here, the Hadamard transform coefficients h0,0 (x, y) are all black as shown in FIG. 6A, and are all “1” as transform coefficients, and all sub-pixel block difference values are added. Corresponding to the difference value between the average value of the pixel values of each pixel of the original image block and the DC value.
Since the sum of all the sub-pixel block difference values is equal to “0” except for the error, the numerical value itself is not stored in the D-RAPIC encoding unit 2.

Each diagram of n = 0 to 3 shown in FIG. 6B is a combination of Hadamard transform coefficients corresponding to each sub-pixel block (2 × 2 pixels) in the pixel block.
For example, in the pixel block S of FIG. 4B, when the correspondence relationship of the conversion coefficient in the equation (8) for each pixel is confirmed, k0 (u, v) (x, y) at n = 0 (sub-pixel block S1). ), U, v, x, y ∈ [0, 1], and when n = 1 (sub-pixel block S2), k 1 (u, v) (x, y), u, v, x, y ∈ [0, 1], and n = 2 (subpixel block S3), k2 (u, v) (x, y), u, v, x, y∈ [0,1], and n = 3 (subpixel block S4) K3 (u, v) (x, y), where u, v, x, y∈ [0,1], and when subscript “n” is removed, in each sub-pixel block, FIG. This is the same as the correspondence of A1 to A4.
Here, Hadamard transform coefficients k0 (0,0) (x, y), k1 (0,0) (x, y), k2 (0,0) (x, y), k3 (0,0) (x , y), x, y∈ [0, 1], as shown in FIG. 6B, all are black, all are “1” as conversion coefficients, and the difference values of all pixels In other words, corresponding to the difference value between the average value of the pixel values in each sub-pixel block and the new first DC pixel value.
Since the sum of the difference values of all the pixels is equal to “0” except for the error, the numerical value itself is not stored in the encoding unit 2.

In the above equation (8), the difference values corresponding to the pixels d (0,0) to d (3,3) are made to correspond to the codes of the corresponding pixels, and the difference values en (0,0) to en ( 1, 1) (this correspondence is associated by comparing the block S in FIG. 4 (e) and FIG. 5 (b)).
Here, the second-stage Hadamard encoding unit 107 calculates the Hadamard coefficient corresponding to each of the sub-pixel blocks S1, S2, S3, and S4 by using the gn already shown in the equation (7) and the equation (8). I do.
The subscript “n” of gn also corresponds to the position of each sub-pixel block (S1 to S4), as in FIG.

In addition, the second-stage Hadamard encoding unit 107 calculates the 12 Hadamard coefficients of each sub-pixel block (for example, S1, S2, S3, S4) included in the pixel block (for example, S) by the equation (8), That is, βn (0,1), βn (1,0), βn (1,1), where n∈ [0,3] is calculated as the second Hadamard coefficient, is quantized, and is output. .
For example, the second-stage Hadamard encoding unit 107 performs inverse quantization on the quantized second Hadamard coefficient, reproduces the second difference value, and corresponds the reproduced second difference value to each part. I is decoded by the D-RPIC decoding unit 13 decoding unit 2 to be described later and output as an image after applying the second stage Hdm (Hadamard transform) (2f in FIG. 3). The same image as the picture is reproduced and stored in the picture storage memory 3.

<Inter coding unit 5>
Next, FIG. 7 is a block diagram illustrating a configuration example of the inter coding unit 5 according to an embodiment of the present invention.
The inter coding unit 5 compares the pixel values of corresponding pixels between the coding target block of each picture (macroblock of 8 × 8 pixels) and the reference block of the reference image (macroblock of 8 × 8 pixels). In order to perform Hadamard transform, a sub-block 201 that takes the difference and outputs it as a residual is divided into a macro block (8 × 8 pixels), a pixel block (4 × 4 pixels), and a sub-pixel block (2 × 2 pixels) Is generated, that is, a pixel block (4 × 4 pixels) that is a processing unit for obtaining the first Hadamard coefficient that is a low-frequency component is decomposed into generated (divided) signals. A macroblock separation unit 202 that generates (divides) a sub-pixel block (2 × 2 pixels) that is a processing unit for obtaining a Hadamard coefficient of 2, and a DC value generation unit 20 that generates a DC value of the pixel block A first-stage Hadamard encoding unit 204 that generates a first Hadamard coefficient, a second-stage Hadamard encoding unit 205 that generates a second Hadamard coefficient, a DC value (DC coefficient), and a first Hadamard coefficient And a macroblock synthesizing unit 207 that synthesizes (combines) the pixel blocks (decoded sub-residual pixel blocks) synthesized by the second Hadamard coefficient so as to correspond to the positions in the macroblocks before the division. And an adder 208 that adds pixel values at corresponding positions of the macro block and the reference block.

The subtraction unit 201 receives an encoding target block and a reference block which is extracted from the reference image by the motion estimation unit and minimizes the residual from the encoding target block, and the encoding target block is input from the reference block. To obtain a residual macroblock that is a difference between the reference block and the encoding target block (residual pixel value of a difference between corresponding pixel values).
The macroblock separation unit 202 divides a residual macroblock composed of 8 × 8 pixels into residual blocks in units of 4 × 4 pixel blocks for low-frequency component extraction in order to perform Hadamard transform in the subsequent stage, Further, it is divided into sub-residual pixel blocks in units of 2 × 2 pixels for extracting high-frequency components.
As shown in FIG. 8, the macro block is composed of four 4 × 4 pixel block, and the pixel block is composed of four 2 × 2 pixel sub-pixel blocks.

The DC value creating unit 203 calculates a residual DC value for each residual pixel block, that is, an average value of pixel values (residual pixel values) included in the residual pixel block for each residual pixel block. The first stage Hadamard encoding unit 204 of the next stage is output to the variable length encoding unit 7 after quantization of the residual DC value.
For each residual pixel block, the first-stage Hadamard encoding unit 204 generates a first difference value (corresponding to the residual pixel block and the residual DC value) from the residual DC value and the residual pixel block. The difference value of the residual pixel value for each pixel) is subtracted, and the first Hadamard coefficient is calculated and output with respect to the first difference value by a basis vector having a predetermined low sequence. The residual pixel block is decoded using one Hadamard coefficient and the residual DC value, that is, the first residual value decoded from the first Hadamard coefficient and the residual DC value are added to obtain a decoded residual. It outputs as a difference pixel block (block after 1st step Hadamard application).
Here, the first stage Hadamard encoding unit 204 performs the same processing as the first stage Hadamard encoding unit 5 in the D-RAPIC encoding unit 2 using, for example, the basis vectors shown in FIG. Then, the first Hadamard coefficient is calculated, and the obtained first Hadamard coefficient is output to the variable length coding unit 7.

  For each sub residual pixel block, the second-stage Hadamard encoding unit 205 calculates the second difference value between the decoded residual pixel block and the residual pixel block, that is, the second residual value between the residual pixel block and the decoded residual pixel block. (Difference value of corresponding pixel value for each pixel) is subtracted, and the second Hadamard coefficient is calculated and output with respect to the second difference value by a predetermined high-basis vector, and this The sub residual pixel block is decoded using the second Hadamard coefficient and the decoded residual pixel block, that is, the second difference value decoded from the second Hadamard coefficient and the decoded residual pixel block are added. Then, it outputs as a decoded sub-residual pixel block (block after the second stage Hadamard application).

The second-stage Hadamard encoding unit 205 performs quantization after collectively quantizing the second Hadamard coefficient and the first Hadamard coefficient generated by the first-stage Hadamard encoding unit 204. The converted result is output to the variable length coding unit 7.
Here, the second stage Hadamard encoding unit 205 performs the same processing as the second stage Hadamard encoding unit 7 in the D-RAPIC encoding unit 2 using, for example, the basis vectors shown in FIG. Then, the second Hadamard coefficient is calculated, and the obtained second Hadamard coefficient is output to the variable length coding unit 7.

  The macroblock synthesis unit 207 synthesizes the decoded sub residual pixel block in units of macroblocks so as to correspond to the position of the macroblock before division by the macroblock separation unit 202, and outputs the result as a decoded residual macroblock. The adder 208 adds the decoded residual macroblock and the reference macroblock (adds the pixel values of the pixels at the corresponding positions, respectively), and adds the addition result to the picture storage memory 3 as a decoded macroblock. , And stored in correspondence with the picture.

<Intra Coding Unit 6>
Next, FIG. 9 is a block diagram illustrating a configuration example of the intra encoder 6 according to an embodiment of the present invention.
The DC image creation unit 301 divides the encoding target block determined to be an intra block in the original image into pixel block units of 4 × 4 pixels, and calculates the average of the pixel values of the pixels included in each pixel block. Thus, a DC value is obtained to generate a DC image, and after the DC value is quantized, the DC value is output to the first stage inter partial complementing unit 302 and also output to the variable length encoding unit 7.
Since the first-stage inter partial complementing unit 302 reproduces the intra block by the ACP method, when the necessary adjacent macroblock is an inter block (see FIG. 8), the first stage inter partial complementing unit 302 uses the picture storage memory 3 to the inter coding unit 5 All the adjacent inter blocks are called from the decoded macro block (inter block) decoded by, and divided into 4 × 4 pixel block units, and the DC value of the pixel block adjacent to the intra block is calculated. The value is output to the first stage AC component prediction unit 303.

The first-stage AC component prediction unit 303 performs the same operation as the first-stage AC component prediction unit 104 of the D-RAPIC encoding unit 2 in the relationship of FIG. DC pixel values are obtained and output as the first-stage AC component prediction applied image.
The first stage Hadamard encoding unit 304 performs the same operation as the first stage Hadamard encoding unit 105 of the D-RAPIC encoding unit 2, and uses the first DC pixel value of the sub-pixel block of the intra block of the original image. Then, the DC value of each pixel block at the corresponding position in the intra block is subtracted to calculate a first difference value that is a difference between the first DC pixel value and the DC value, and the first difference value is calculated with respect to the first difference value. Then, the first Hadamard coefficient is obtained by applying the basis vector of FIG.
Further, the first stage Hadamard encoding unit 304 decodes the first DC pixel value from the DC value and the first Hadamard coefficient, and the first decoded DC value (the first stage Hadamard applied image: intra Output to the second-stage inter-partial complementing unit 305 as image data in block units).

The second-stage inter partial complementing unit 305 calls all adjacent inter blocks from the decoded macroblock (interblock) decoded by the inter coding unit 5 from the picture storage memory 3, and subpixel blocks adjacent to the intra block The average of the pixel values of the included pixels is obtained and output as the first pixel value.
The second-stage AC component prediction unit 306 performs the same operation as the second-stage AC component prediction unit 106 of the D-RAPIC encoding unit 2, and the first decoded DC value and the first block obtained from the intra block. The second DC pixel value of each sub-pixel block is decoded and obtained from the pixel value, and the second DC pixel value is obtained by applying the second-stage AC component prediction image (image data of intra block unit). To the second-stage Hadamard encoding unit 307.

The second stage Hadamard encoding unit 307 performs the same operation as that of the second stage Hadamard encoding unit 107 of the D-RAPIC encoding unit 2, and the pixel value of each pixel of the sub-pixel block corresponding to the macroblock of the original image From the above, the second DC pixel value is subtracted to obtain a second difference value that is the difference between the pixel value and the second DC pixel value. ) To obtain the second Hadamard coefficient.
Further, the second stage Hadamard encoding unit 307 quantizes the second Hadamard coefficient and the first Hadamard coefficient generated by the first stage Hadamard encoding unit 304, and then performs intra coding. The encoded data is output to the variable length encoding unit 7.

  Further, the second-stage Hadamard encoding unit 307 decodes and determines the pixel value of each pixel in the sub-pixel block based on the second DC pixel value and the second Hadamard coefficient, and obtains this decoded pixel value as the second pixel value. It is stored in the picture storage memory 3 as an image after application of stage Hadamard (image data in units of intra blocks).

<< Decoding Unit >>
Next, the image processing apparatus of the present invention will be described with reference to the drawings. FIG. 10 is a block diagram illustrating a configuration example of an image processing apparatus that performs decoding according to an embodiment of the present invention.
The image processing apparatus includes a determination unit 10, a variable length decoding unit 11, a variable length decoding unit 12, a D-RAPIC decoding unit 13, an inter decoding unit 14, an intra decoding unit 15, a picture storage memory 16, and a picture rearrangement unit 17. Have.
The determination unit 10 receives the encoded data from the encoding unit already described, and determines whether the input encoded data is an I picture or a P or B picture.

If the determination unit 10 determines that the input encoded data is an I picture, the determination unit 10 outputs the data to the variable length decoding unit 11. If the determination unit 10 determines that the input encoded data is a P or B picture, the determination unit 10 outputs the data to the variable length decoding unit 12. To do.
The variable length decoding unit 11 decodes the encoded data, and outputs it to the D-RAPIC decoding unit 13 after decoding as a DC value, a first Hadamard coefficient, and a second Hadamard coefficient.

The variable length decoding unit 12 converts the input encoded data into DC pixel value, first Hadamard coefficient, second Hadamard coefficient, MCB encoded information, and motion vector information as decoded data after the decoding. The data is output to one of the intra decoders 15.
At this time, the variable length decoding unit 12 determines whether each input encoded data is intra-coded or inter-coded based on the MCB coding information, and is inter-coded. If it is determined, the DC value (residual DC value), the first and second Hadamard coefficients, and the motion vector information are output to the inter decoding unit 14.
On the other hand, the variable length decoding unit 12 outputs the DC value and the first and second Hadamard coefficients to the intra decoding unit 15 when it is determined that each piece of input encoded data is intra-encoded.

As will be described in detail later, the D-RAPIC decoding unit 13 decodes the I picture by calculation based on the ACP (alternating current component prediction) method, and writes the decoded data into the picture storage memory 16.
In the picture storage memory 16, the I picture decoded by the D-RAPIC decoding unit is written, and the decoded P and B pictures are written by the inter decoding unit 14 and the intra decoding unit 15, respectively. ing.
The picture rearrangement unit 17 encodes the encoded permutations [0 (I), 3 (P), 1 (B), 2 (B), 6 (P), 4 converted by the picture rearrangement unit 1 as shown in FIG. (B), 5 (B), 9 (I), 7 (B), 8 (B), 12 (B), 10 (B), 11 (B)] are read from the picture storage memory 16 Image permutation [0 (I), 1 (B), 2 (B), 3 (P), 4 (B), 5 (B), 6 (P), 7 (B), 8 (B), 9 (I), 10 (B), 11 (B), 12 (P)] are rearranged (decoded) and sequentially output in time series to reproduce a moving image.

The inter decoding unit 14 decodes inter-coded P and B pictures based on motion vector information, a DC value (residual DC value), a first Hadamard coefficient, a second Hadamard coefficient, MCB coding information, and the like. I do.
Similarly, the intra decoding unit 15 performs intra-coded P and P based on the macroblock decoded by the inter decoding unit 14, the DC value, the first Hadamard coefficient, the second Hadamard coefficient, MCB encoded information, and the like. Decode the B picture.

Hereinafter, the configuration and operation examples of the D-RAPIC decoding unit 13, the inter decoding unit 14, and the intra decoding unit 15 according to the embodiment of the present invention will be sequentially described with reference to the drawings.
<D-RAPIC decoding unit 13>
First, the outline of the configuration and operation of the D-RAPIC decoding unit 13 will be described.
Similar to the D-RAPIC encoding unit 2 described above, the D-RAPIC decoding unit 13 according to the present invention performs a first step ACP process based on the DC value of the input DC image during decoding, A DC pixel value is obtained.
Next, the D-RAPIC decoding unit 13 inversely quantizes the input first Hadamard coefficient, performs first-stage Hadamard decoding (inverse Hadamard transform), obtains a first difference value, and the first difference The value is added to the first DC pixel value of the corresponding sub-pixel block to generate a new first DC pixel value.

Next, the D-RAPIC decoding unit 13 performs a second stage ACP process using the first DC pixel value, and obtains a second DC pixel value for each pixel.
Then, the D-RAPIC decoding unit 13 performs the second-stage Hadamard decoding by dequantizing the input second Hadamard coefficient, and obtains the second difference value. The difference value is added to the second DC pixel value of each corresponding pixel to generate a new second DC pixel value, which is output as a reproduction pixel value.

  As described above, since the D-RAPIC decoding unit 13 includes not only Hadamard transform but also component prediction processing from surrounding pixel blocks or sub-pixel blocks by the ACP method, not only mosquito noise but also Hadamard transform is blocked. Block noise that tends to appear because it is performed in units can also be reduced.

  In addition, the D-RAPIC decoding unit 13 can realize the processing described above by performing only the operations of addition and subtraction and bit shift for encoding and decoding, and thus can be realized with a compact processing configuration. Since there is no branch processing associated with the difference in size and processing can be performed continuously at a predetermined cycle, high-speed encoding and decoding (inverse operation of encoding) can be performed.

Next, FIG. 11 is a block diagram showing a configuration example of the D-RAPIC decoding unit 13 according to an embodiment of the present invention.
The DC image decoding unit 151 performs inverse quantization and decoding on the input DC code to obtain a DC value, and reproduces a DC image (4b in FIG. 11).
Similarly to the first-stage AC component prediction unit 104, the first-stage AC component prediction unit 152 obtains the first DC pixel values of the sub-pixel blocks S1 to S4, and the first-stage AC component prediction unit 152 includes the first-stage AC component prediction unit 152. It outputs as a step ACP application image (4c of FIG. 11).
The first stage Hadamard decoding unit 153 performs inverse quantization and entropy decoding on the input encoded first Hadamard coefficient, and decodes (reproduces) it as Hadamard coefficients α′01, α′10, α′11, The first Hadamard coefficient is decoded (reproduced) by performing inverse Hadamard transform using the following equation (9).

ここで、アダマール変換係数ｈ0,0(x,y)に対応するα'00は、前述の通り「０」に等しいため、D-RAPIC復号部１３内部に保存されない。
そして、（９）式のｆ'は下記に示す（１１）式として求められる。

That is, the first-stage Hadamard decoding unit 153 calculates and calculates the first difference value corresponding to each part of the sub-pixel block to be reproduced using the following equation (10).

Here, α′00 corresponding to the Hadamard transform coefficient h0,0 (x, y) is equal to “0” as described above, and thus is not stored in the D-RAPIC decoding unit 13.
And f 'of (9) Formula is calculated | required as (11) Formula shown below.

  In addition, the first-stage Hadamard decoding unit 153 generates a decoded first difference value for each first DC pixel value of each of the sub-pixel blocks S1 to S4 input from the first-stage AC component prediction unit 152. , By adding for each corresponding sub-pixel block, the addition result becomes a new first DC pixel value of each of the sub-pixel blocks S1 to S4, and this sub-pixel block S1 to S4 is the image after applying the first stage Hdm ( It is output as 4d) in FIG.

すなわち、以下に示す（１３）式により、各サブ画素ブロック毎に、第２段階アダマール符号部１０７と同様に、図４（ｆ）及び（７）,（８）式の関係に対応させて、サブ画素ブロックにおける各画素に対応する第２の差分値を演算する。

そして、アダマール変換係数ｋ0（0,0)(x,y)，ｋ1（0,0)(x,y)，ｋ2（0,0)(x,y)，ｋ3（0,0)(x,y)、x,y∈[0,1]各々に対応するβ'n(0,0)，ｎ∈[0,3]は、前述の通り「０」に等しいため、D-RAPIC復号部１３内部に保存されない。 Next, the second-stage AC component prediction unit 156 performs the second-stage AC based on the first-stage Hdm-applied image (4d in FIG. 11) of each sub-pixel block in the input first-stage Hdm-applied image 4d. A second DC pixel value for each pixel of the sub-pixel block is obtained by calculation using a formula similar to that of the component prediction unit 106, and is output as a second-stage ACP-applied image (4e in FIG. 11).
The second-stage Hadamard decoding unit 157 obtains a difference value for each pixel based on the following expression (12) using the input second Hadamard coefficient.

That is, according to the following equation (13), for each sub-pixel block, as in the second-stage Hadamard encoding unit 107, corresponding to the relationship of equations (f), (7), and (8), A second difference value corresponding to each pixel in the sub-pixel block is calculated.

The Hadamard transform coefficients k0 (0,0) (x, y), k1 (0,0) (x, y), k2 (0,0) (x, y), k3 (0,0) (x, Since β′n (0,0) and n∈ [0,3] corresponding to each of y), x, y∈ [0,1] are equal to “0” as described above, the D-RAPIC decoding unit 13 Not stored internally.

Here, the second-stage Hadamard decoding unit 157 receives the second Hadamard coefficient βn (u, v), where n∈ [0,3], u, v∈ [0, 1] and a Hadamard transform coefficient, the second difference value corresponding to each pixel is calculated.
The Hadamard transform coefficient for obtaining the pixel of each sub-pixel block is k0 (u, v) (x, y), u, v, x, y∈ [0,1] at n = 0 (sub-pixel block S1). N = 1 (subpixel block S2), k1 (u, v) (x, y), u, v, x, y∈ [0,1], and n = 2 (subpixel block S3) K2 (u, v) (x, y), u, v, x, y∈ [0,1], and k = 3 (u, v) (x, y) at n = 3 (sub-pixel block S4) , U, v, x, y∈ [0, 1].
Further, the second stage Hadamard decoding unit 157 adds the second difference value obtained for each pixel unit to the second DC pixel value of the pixel at the corresponding position, and then applies the second stage Hdm applied image ( As a new second DC pixel value of the sub-pixel block in 4q) of FIG. 11, that is, as an I picture that is a decoded image, it is output to the picture storage memory 16 and stored.

このｇ'nの添え字「ｎ」も、図４（ｅ）と同様に、各サブ画素ブロック（S1〜S4）の位置に対応している。 Thereby, g′n in the equation (12) is obtained as the following equation (14).

The subscript “n” of g′n also corresponds to the position of each sub-pixel block (S1 to S4), as in FIG.

As described above, the image processing apparatus according to the present invention has a low frequency component (first frequency) with a large amount of information in the first Hadamard transform among the 15 Hadamard coefficients output as frequency components of 16 pixels. After extracting three of the Hadamard coefficients), this is decoded and added to the first DC pixel value to obtain the second DC pixel value of each pixel of the sub-pixel block, and then the second Hadamard transform performs high frequency Three components (second Hadamard coefficients) are obtained for each of the four sub-pixel blocks, and a total of 12 Hadamard coefficients are obtained.
For this reason, the data amount of the second Hadamard coefficient can be reduced as compared with the one-step process in which the ACP process and the Hadamard transform are performed once.

<Inter decoding unit 14>
FIG. 12 is a block diagram showing a configuration example of the inter decoding unit 14 according to an embodiment of the present invention.
The first-stage Hadamard decoding unit 402 has a DC value (residual DC value) in units of pixel blocks of the input residual macroblock, and the first and second Hadamard coefficients (for the macroblock to be decoded). The input DC value and the first and second Hadamard coefficients are four (because there are 4 × 4 pixel sub-pixel blocks in one macroblock), so that the first-stage Hadamard in the D-RAPIC decoding unit 13 The first difference value for each pixel block is obtained by the same arithmetic processing as the decoding unit 153, and for each pixel block, the first difference value is added to the pixel value for each corresponding sub-pixel block, and the addition result Are output to the second-stage Hadamard decoding unit 403 as blocks after the first-stage Hadamard application.

  The second-stage Hadamard decoding unit 403 divides the input residual pixel block into sub-pixel block units, applies a second Hadamard coefficient for each sub-pixel block (residual sub-pixel block), and first-stage Hadamard application. With the subsequent block, the second difference value in units of pixels in the sub-pixel block is obtained by the same arithmetic processing as in the second-stage Hadamard decoding unit 157 in the D-RAPIC decoding unit 13, and each pixel corresponding to each sub-pixel block is obtained. Then, the second difference value is added, and the addition result is output to the macroblock synthesis unit 404 as the second-stage Hadamard applied block.

The macroblock synthesis unit 404 is a 2 × 2 synthesized by the second-stage Hadamard decoding unit 403 using the DC value (DC coefficient), the first Hadamard coefficient, and the second Hadamard coefficient decoded for each sub-pixel block. A pixel block (decoded sub-residual pixel block) of pixels is synthesized (combined) so as to correspond to a position in a macroblock of 8 × 8 pixels before division (before separation).
The adder 405 adds the difference value (residual pixel value) and the pixel value in each of the corresponding pixels of the synthesized residual macroblock and the reference block, and the macro in which the encoding target block is decoded The block-based picture, that is, a P or B picture is output to the picture storage memory 16 and stored.

<Intra decoding unit 15>
FIG. 13 is a block diagram illustrating a configuration example of the intra decoding unit 15 according to an embodiment of the present invention.
The first stage inter partial interpolation unit 502 outputs the DC value of each 4 × 4 pixel block in the input macroblock to the first stage AC prediction component predictor 503 and is adjacent to the macroblock to be decoded. The pixel value of the macroblock subjected to the inter coding is read out from the picture storage memory 16, and, similar to the first stage inter partial interpolation unit 302 of the intra coding unit 6, for each pixel block in the adjacent macroblock, The average of the included pixel values is calculated to determine the DC value of each pixel block in the inter-decoded macroblock.

  The first-stage AC component prediction unit 503 uses the DC pixel value decoded by the variable-length code decoding unit 12 and the DC pixel value obtained by each of the first-stage inter-partial interpolation unit 502, and uses the D-RAPIC decoding unit 13 Similar to the first-stage AC component prediction unit 152, the first DC pixel value for each sub-pixel block is calculated by the ACP method, and is output to the first-stage Hadamard decoding unit 504 as the first-stage AC component prediction applied image. .

  The first-stage Hadamard decoding unit 504 performs the same calculation process as that of the first-stage Hadamard decoding unit 153 in the D-RAPIC decoding unit 13 based on the input first Hadamard coefficient for each pixel block. 1 difference value is obtained, and for each pixel block, this first difference value is added to the pixel value for each corresponding sub-pixel block, and the addition result is used as a block after first-stage Hadamard application and second-stage inter-partial interpolation Output to the unit 505.

  The second-stage inter-part interpolation unit 505 reads the pixel value of the macroblock adjacent to the macroblock to be decoded and subjected to inter-coding from the picture storage memory 16, and the second-stage inter-part of the intra-encoding unit 6 Similar to the interpolation unit 302, the average of the included pixel values is calculated for each sub-pixel block in each pixel block of the adjacent macro block, and the first of the sub-pixel blocks in each pixel block of the inter-decoded macro block is calculated. Are obtained and output to the second-stage AC component prediction unit 506.

  The second-stage AC component prediction unit 506 uses the first DC pixel values obtained by the first-stage AC component prediction unit 503 and the second-stage inter partial interpolation unit 505, and uses the second DC component value in the D-RAPIC decoding unit 13. Similar to the step AC component prediction unit 306, the second DC pixel value of each pixel in the sub-pixel block is calculated by the ACP method, and is output to the second step Hadamard decoding unit 507 as the second step AC component prediction applied image. .

  The second-stage Hadamard decoding unit 507 performs the same calculation process as that of the second-stage Hadamard decoding unit 157 in the D-RAPIC decoding unit 13 on the basis of the second Hadamard coefficient for each input pixel block. A second difference value in pixel units is obtained, and for each pixel block, the second difference value is added to the pixel value for each pixel in the corresponding sub-pixel block, and the addition result is a block after the second-stage Hadamard application, That is, the decoded image is output to the picture storage memory 16 and stored.

The above-described image compression and decoding processes are performed using image data of Y, U, and V (luminance = Y, red color difference = U, blue color difference = V) used for digital moving images.
For RGB (Red = red, Green = green, Blue = blue) image data, it is necessary to convert the image data into YUV format image data.
The image processing apparatus of the present invention performs similar processing on Y, U, and V, and compresses and decodes images.

As described above, the image processing apparatus according to the present invention has 15 Hadamard coefficients output as frequency components of 16 pixels in the D-RAPIC encoding unit 2, the inter encoding unit 5, and the intra encoding unit 6. In the first Hadamard transform, after extracting three low frequency components having a large amount of information, they are decoded and added to the first DC pixel value, and then the second DC of each pixel of the sub-pixel block. After obtaining the pixel values, three high-frequency components are obtained for each of the four sub-pixel blocks in the second Hadamard transform, for a total of 12 Hadamard coefficients.
For this reason, the data amount of the second Hadamard coefficient can be reduced as compared with the one-step process in which the ACP process and the Hadamard transform are performed once.

  Further, in the image processing apparatus according to the present invention, when the encoding target macroblock is encoded in the intra encoding unit 6, the macroblock of the inter-coded portion adjacent to the encoding target macroblock to be intra-encoded is used. Then, the DC value and the first DC pixel value of each of the pixel block and the sub-pixel block are obtained, and the first and second difference values are obtained using these, and the DC value used for the ACP process and the Hadamard transformed difference are obtained. Therefore, it is possible to perform ACP processing of a macroblock to be intra-coded, to reduce the amount of information necessary for image reproduction processing, and to perform intra-coded macroblock and inter-coding. The encoding efficiency at the boundary with the converted macroblock can be improved.

  In addition, in the image processing apparatus according to the present invention, the D-RAPIC encoding unit 2 and the intra encoding unit 6 perform the DC value for performing the ACP process, and the first value after the ACP process is performed on the DC value. The difference value between the first and second DC pixel values and the original image is output as compression information as a Hadamard coefficient, and the inter encoding unit 5 uses the difference value between the reference image data and the encoding target block as a Hadamard coefficient. Since it is output as compressed information, the amount of information required for image reproduction processing is reduced, and above all, the difference value from the original image is output as a Hadamard coefficient, so that the original image can have high reproducibility. Is possible.

  Further, in the image processing apparatus according to the present invention, in the D-RAPIC encoding unit 2 and the intra encoding unit 6, not only the Hadamard transform described above but also the first-stage AC component prediction units 104 and 303 and the second-stage AC component, respectively. Since image processing based on the ACP method is also added in the prediction units 106 and 306, a rapid change in gradation between adjacent pixel blocks or sub-pixel blocks, that is, a change in high spatial frequency in the image is reduced. Since the block noise of the image after the decoding process can be reduced, the entire image can be decoded smoothly and naturally, and the block noise resulting from the Hadamard transform encoding for the pixel block or sub-pixel block can be reduced.

  Furthermore, the image processing apparatus according to the present invention includes the first-stage AC component prediction units 104, 152, 303, and 503, the second-stage AC component prediction units 106, 156, 306, and 506, and the first-stage Hadamard encoding unit 105. , 204, 304, second-stage Hadamard encoding sections 107, 205, 307, first-stage Hadamard decoding sections 153, 402, 504, and second-stage Hadamard decoding sections 157, 403, 507, respectively, 8), (10), and (13) can be used to perform the above-described processing only by addition and subtraction and bit shift operations, so that a compact circuit configuration can be realized. Since there is no branch processing in compression or expansion accompanying the magnitude of the value, and processing can be performed continuously in a predetermined cycle, high-speed encoding and decoding processing can be performed.

  In addition, in the image processing apparatus according to the present invention, in the encoding unit and the decoding unit, the first-stage AC component prediction unit (for example, 104 and 303 are shared or 152 and 503 are shared) for each unit, And the second-stage AC component prediction unit (for example, 106 and 306 are shared and 156 and 506 are shared), and similarly, the first-stage Hadamard encoding unit and the second-stage Hadamard Since the encoding unit, the first stage Hadamard decoding unit, and the second stage Hadamard decoding unit can be shared, it is possible to improve the compactness of the processing configuration.

Here, since the Hadamard transform coefficient is “1” or “−1”, the product of the Hadamard transform coefficient with respect to the pixel value and the Hadamard coefficient in the expressions (3), (8), (10), and (13). Is calculated as a sign addition rather than a multiplication.
In other operations, the basic operation is addition as can be seen from the equation, and the quantization is performed by bit shift.

Next, the results of comparison between the case of using the compression and decoding method of the present invention and the case of using MPEG2 will be described.
FIG. 14 is a graph showing the degree of image quality with respect to the compression rate, with the horizontal axis indicating the bit amount per pixel (BPP: Bit per Pixel) and the vertical axis indicating PNSR (peak value to error ratio).
Here, the compression rate is the ratio of the number of bytes before and after compression of the image file, and the lower the value, the higher the compression rate.
The solid line indicates the relationship between the file size and PNSR when the image processing apparatus of the present invention is used, and the broken line indicates the relationship between the file size and PNSR when MPEG2 is used.

From the comparison between the image processing apparatus of the present invention and MPEG2, when decoding is performed from the same compression rate, in the present invention, the numerical image quality is improved by PNSR in the region where the number of bits of the file is high.
The above-described image data is processed at a ratio of “Y: U: V = 4: 1: 1” as the number of processed pixels in both the present invention and MPEG2.
An image used as a sample has 8-bit color information for each RGB in one pixel.

  As mentioned above, although one embodiment of the present invention has been described in detail with reference to the drawings, the specific configuration is not limited to this embodiment, and there are design changes and the like without departing from the gist of the present invention. Are also included in the present invention.

Next, a program executed by the computer according to the embodiment of the present invention will be described.
The program executed by the CPU of the computer system in the operation of the image processing apparatus (image processing system) in FIGS. 1 and 10 constitutes a program according to the present invention.
As a recording medium for storing this program, a magneto-optical disk, an optical disk, a semiconductor memory, a magnetic recording medium, and the like can be used, and these are configured as a ROM, a RAM, a CD-ROM, a flexible disk, a memory card, and the like. May be used.

In addition, the recording medium can store a program for a certain period of time, such as a volatile memory such as a RAM in a computer system serving as a server or a client when the program is transmitted via a network such as the Internet or a communication line such as a telephone line. The thing to hold is also included.
The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. The transmission medium refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.

  The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

  Therefore, even when this program is used in a system or apparatus different from the system or apparatus shown in FIGS. 1 and 10 and the computer of the system or apparatus executes the program, the functions and effects described in the above embodiments can be obtained. Equivalent functions and effects can be obtained, and the object of the present invention can be achieved.

It is a block diagram which shows the structural example of the encoding part of the image processing apparatus by one Embodiment of this invention. It is a conceptual diagram explaining operation | movement of the picture rearrangement parts 1 and 17 by one Embodiment of this invention. It is a block diagram which shows the structural example of the D-RAPIC encoding part 2 by one Embodiment of this invention. It is a conceptual diagram for demonstrating the operation example of the D-RAPIC encoding part 2 shown in FIG. It is a conceptual diagram explaining the image processing by ACP method of the 1st step and the 2nd step. It is a conceptual diagram which shows an example of the Hadamard transform coefficient used in Hadamard transform. It is a block diagram explaining the structural example of the inter encoding part 5 in the encoding part of FIG. It is a conceptual diagram which shows the correspondence of a macroblock and a sub pixel block. It is a block diagram which shows the structural example of the intra encoding part 6 in the encoding part of FIG. It is a block diagram which shows the structural example of the decoding part of the image processing apparatus by one Embodiment of this invention. It is a block diagram which shows the structural example of the D-RAPIC decoding part 13 by one Embodiment of this invention. It is a block diagram which shows the structural example of the inter decoding part 14 by one Embodiment. It is a block diagram which shows the structural example of the intra decoding part 15 by one Embodiment. It is a graph which shows the relationship between the file size and PNSR in the image processing apparatus of this invention and the process of compression and decoding by MPEG2.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1,17 Picture rearrangement part 2 D-RAPIC encoding part 3,16 Picture storage memory 4 Motion estimation part 5 Inter encoding part 6 Intra encoding part 7 Variable length encoding part 10 Judgment part 11, 12 Variable length decoding part 13 D-RAPIC decoding unit 14 Inter decoding unit 15 Intra decoding unit 103, 301 DC image creation unit 104, 152, 303, 503 First stage AC component prediction unit 105, 204, 304 First stage Hadamard encoding unit 106, 156 , 306, 506 Second-stage AC component prediction section 107, 205, 307 Second-stage Hadamard encoding section 151 DC image decoding sections 153, 402, 504 First-stage Hadamard decoding sections 157, 403, 507 Second-stage Hadamard decoding section 201 Subtraction unit 202 Macroblock separation unit 203 DC value creation unit 207, 404 Macroblock synthesis 208, 405 Adder 302, 502 First stage inter partial interpolator 305, 505 Second stage inter partial interpolator

Claims (6)

In an image processing apparatus for encoding a moving image using an I picture, a P picture, and a B picture,
When the input original image is an I picture, the original image is divided into pixel blocks of 4 × 4 pixel units, a DC value that is an average value of the pixel values is obtained, and an AC component between adjacent pixel blocks For each 2 × 2 pixel sub-pixel block obtained by performing prediction and dividing each pixel block into four, a first DC pixel value is obtained, and the first DC pixel value and the sub-pixel block of the original image at the corresponding position Is obtained as a first difference value, the first difference value is encoded as a first Hadamard coefficient, and the DC value and the first Hadamard coefficient are output as encoded data. An encoding unit;
The pixel block is divided into sub-pixel blocks of 2 × 2 pixel units, a DC value that is an average value of pixel values is obtained, AC component prediction is performed between adjacent sub-pixel blocks, and each pixel in each sub-pixel block is determined. Then, a second DC pixel value is obtained, a difference between the second DC pixel value and a pixel in the sub-pixel block of the original image at the corresponding position is obtained as a second difference value, and the second difference value is obtained. Is encoded as a second Hadamard coefficient, and the second AC prediction encoding unit that outputs the second Hadamard coefficient as encoded data;
When the input original image is a B or P picture, the original image is divided into macro block units of 8 × 8 pixels, and a reference block having the highest correlation is extracted from a predetermined other picture. Using a block, a motion estimation unit for determining whether the encoded data for motion compensation of the macroblock is within a predetermined data capacity range;
When the coded data for motion compensation of the macroblock is within the predetermined range, an inter coding unit that performs inter coding based on motion compensation for the macroblock;
An image processing apparatus comprising: an intra coding unit that performs intra coding on a macroblock when coded data for motion compensation of a macroblock is not within the predetermined range.   The inter coding unit obtains a difference value between an 8 × 8 pixel reference image (reference block) extracted by motion prediction and a macroblock to be coded, and the difference block based on the difference value is determined as a 4 × 4 pixel. A block (residual block) is divided into units, and for each pixel block, the difference value is averaged to obtain a DC value, and a difference between the DC value and the difference value of the difference block is encoded as a Hadamard coefficient, The image processing apparatus according to claim 1, wherein the DC value and Hadamard coefficient are output as encoded data. The intra encoding unit is
The macroblock to be encoded is divided into pixel blocks of 4 × 4 pixel units, a DC value that is an average value of pixel values is obtained, AC component prediction is performed between adjacent pixel blocks, and each pixel block is divided into 4 A first DC pixel value is obtained for each divided 2 × 2 pixel sub-pixel block, and a difference between the first DC pixel value and a sub-pixel block at a corresponding position in the macro block is determined as a first difference. A third AC predictive encoding unit that obtains the first difference value as a first Hadamard coefficient and outputs the DC value and the first Hadamard coefficient as encoded data;
The pixel block is divided into sub-pixel blocks of 2 × 2 pixel units, a DC value that is an average value of pixel values is obtained, AC component prediction is performed between adjacent sub-pixel blocks, and each pixel in each sub-pixel block Then, a second DC pixel value is obtained, a difference between the second DC pixel value and a pixel in the sub-pixel block of the original image at the corresponding position is obtained as a second difference value, and the second difference value is obtained. 4. The image processing according to claim 1, further comprising: a fourth AC prediction encoding unit that encodes the second Hadamard coefficient as encoded data and outputs the second Hadamard coefficient as encoded data. apparatus.   When the AC prediction encoding unit divides a macroblock to be encoded into pixel blocks of 4 × 4 pixel units and also subdivides into 2 × 2 pixel sub-pixel blocks and performs AC prediction encoding, 4. The image processing apparatus according to claim 3, wherein when the target pixel block is an inter-coded pixel block, AC component prediction is performed using a macroblock obtained as a result of the decoding. 5. In an image processing apparatus for decoding a moving image using an I picture, a P picture, and a B picture,
When the encoded image data to be input is an I picture, AC component prediction is performed between the DC value of the input target pixel block to be decoded and the DC value of the pixel block adjacent to the target pixel block. The first DC pixel value is obtained for each 2 × 2 pixel sub-pixel block obtained by dividing each pixel block into four, and the inverse of the input first Hadamard coefficient is obtained with respect to the first DC pixel value. Hadamard transform is performed, a difference between the first DC pixel value of the sub-pixel block and the sub-pixel block of the original image at the corresponding position is obtained as a first difference value, and the first difference value is obtained as an AC component. A first AC predictive decoding unit that adds to the first DC pixel value decoded by prediction and outputs the addition result as a new first DC pixel value;
AC component prediction is performed between the first DC pixel value and the first DC pixel value of the sub-pixel block adjacent to the first DC pixel value, and a second DC pixel value is obtained for each pixel. The inverse Hadamard transform of the input second Hadamard coefficient is performed on the second DC pixel value of the pixel, and the second DC pixel value of the pixel and the pixel of the original image at the corresponding position are Second AC predictive decoding that obtains the difference as a second difference value, adds the second difference value to the second DC pixel value decoded by the AC component prediction, and outputs the addition result as decoded data And
A determination unit that determines whether an 8 × 8 pixel macroblock to be decoded is intra-coded or inter-coded when the input encoded image data is a B or P picture; ,
When the macroblock to be decoded is inter-coded, the difference block based on the difference between each pixel of the reference block and the macroblock is divided into 4 × 4 pixel blocks, and the remaining number obtained for each pixel block is divided. In the residual DC value that is the average value of the differences and the first and second Hadamard coefficients obtained for each 4 × 4 pixel block, the residual DC value and the first and second Hadamard coefficients are inverse Hadamard transformed. An inter decoding unit that decodes the macroblock by adding the difference value of the residual obtained in this way and adding the pixel value of the reference block to the addition result;
An intra-decoding unit that decodes a macroblock based on a DC value and a Hadamard coefficient of an input 4 × 4 pixel block when the macroblock to be decoded is intra-coded. An image processing apparatus. When the intra decoding unit divides a macroblock to be decoded into pixel blocks of 4 × 4 pixel units, and further divides into 2 × 2 pixel sub-pixel blocks and performs decoding by AC predictive coding, 6. The image processing apparatus according to claim 5, wherein when the target pixel block is an inter-encoded pixel block, AC component prediction is performed using a macroblock obtained as a result of decoding.

Images