JP2009049969A - Device and method of coding moving image and device and method of decoding moving image - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem that, in the intra-screen predictive coding according to the H.264/AVC, adjacent encoded pixels are used to perform prediction of a target block, so that the prediction needs to be performed by referencing spatially-separate pixels, as such, the prediction accuracy is not always sufficient. <P>SOLUTION: Adjacent pixels are always used to perform the intra-screen predictive coding and decoding. In the coding, the difference value (prediction residue) between the adjacent pixels is calculated in a certain direction (predicting direction) and encoded with the predicting direction. In the coding of each block, adjacent pixels may have not been encoded yet. In that case, the difference value between adjacent pixels in an original image is encoded. On the other hand, in the decoding, processing to add a value produced by decoding the previous pixel value and the prediction residue is repeated in a predicting direction to obtain a decoded image. This can perform highly-efficient coding processing using strong correlation between adjacent pixels. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a moving image encoding technique for encoding and decoding a moving image.

MPEG (Moving) is a method for recording and transmitting large volumes of moving image information as digital data.
Coding schemes such as the Picture Experts Group) were formulated, and the MPEG-1 standard, MPEG-2 standard, MPE
It is an international standard encoding method such as G-4 standard, H.264 / AVC (Advanced Video Coding) standard, and the like. These systems have been adopted as encoding systems for digital satellite broadcasting, DVDs, mobile phones, digital cameras, and the like, and the range of use is now expanding and becoming familiar.

In these standards, the encoding target image is predicted in block units using the image information that has been encoded, and the prediction difference from the original image is encoded, thereby eliminating the redundancy of the moving image. The code amount is reduced. In particular, H.264 / AVC achieved a dramatic improvement in compression rate by adopting an intra-screen predictive encoding method that uses peripheral pixels of the encoding target block.

However, in the case of intra prediction in H.264 / AVC, only pixel information adjacent to the target block can be referred to, and it cannot be said that the prediction accuracy is sufficient. Therefore, an intra-screen coding technique for improving the accuracy of intra-screen prediction and increasing the compression rate has been demanded.
From the background described above, many techniques for improving the accuracy of intra prediction by increasing the types of pixels that can be used for intra prediction are proposed.

For example, in Patent Document 1, the entire image can be inverted before encoding, thereby changing the encoding processing order and increasing the types of pixels that can be used for intra prediction.
Further, in Patent Document 2, it is possible to perform prediction with reference to a distant block by searching for a block used for prediction of a target block from an encoded region in the same screen.

In Patent Document 3, the input block is separated into at least two regions, and intra prediction coding is performed using pixels in the peripheral block for pixels in the first region among the separated regions. Performing the intra-prediction encoded first region pixels, and using the restored first region pixels, the second region pixels may be stored in the second region according to at least one of the plurality of prediction modes. An intra-prediction encoding method including a step of predicting a pixel is disclosed.

Further, in Patent Document 4, after the input block is separated into at least two regions, and intra prediction is performed on the pixels in the first region among the separated regions using the pixels in the peripheral blocks. The spatial characteristics of the pixels in the first area located around the predetermined pixels in the second area of the separated areas are determined, and the pixels in the second area are predicted based on the determined spatial characteristics. A video intra prediction encoding method is disclosed in which a reference pixel of the first region is determined, and a pixel of the second region is predicted based on the determined reference pixel of the first region.

In Non-Patent Document 1, an optimal encoding mode is determined from the relationship between image quality distortion and code amount.
RD-Optimization method is described.
JP 2006-352181 A JP 2007-43651 A JP 2007-74725 A JP 2007-74725 A G. Sullivan and T. Wiegand: Rate-Distortion Optimization for Video Compression ", IEEE Signal Processing Magazine, vol.15, no.6, pp.74-90, 1998.

However, since all the existing intra prediction encoding techniques including the above documents perform prediction using decoded pixels, the intra prediction encoding is performed on a block basis. Therefore, it was necessary to predict with reference to spatially separated pixels. In this case, since the distance between the prediction target pixel and the referenced pixel is increased, there is a problem that the correlation between the pixels is reduced and the prediction accuracy is lowered.

In the intra prediction encoding in H.264 / AVC, since the target block is predicted using the encoded adjacent pixels, it is necessary to perform prediction with reference to spatially separated pixels, and the prediction accuracy Was not enough.

Therefore, the present invention provides a technique for improving prediction accuracy by always performing prediction with reference to adjacent pixels. An object of the present invention is to improve the efficiency of intra prediction encoding by always performing prediction with reference to adjacent pixels having high inter-pixel correlation.

In the present invention, adjacent pixels are always used when performing intra prediction encoding and decoding. At the time of encoding, a difference value (prediction difference) between adjacent pixels is calculated along a certain direction (prediction direction) and encoded together with the prediction direction. At this time, when encoding is performed in block units, adjacent pixels may be in an uncoded state. In this case, a difference value between adjacent pixels in the original image is encoded. On the other hand, at the time of decoding, a decoded image is acquired by repeating the process of adding the prediction difference to the value obtained by decoding the immediately preceding pixel value along the prediction direction. This
A highly efficient encoding process can be performed using a strong correlation between adjacent pixels.

ADVANTAGE OF THE INVENTION According to this invention, the moving image encoding technique and decoding technique for providing a high quality image | video with a small code amount can be provided. The present invention is directed to a next-generation image coding standard. For example, a high compression ratio can be realized by using a predictive coding technique according to the present technique in combination with a conventional technique. For example, by appropriately using the present technology and the intra-screen predictive coding technology based on H.264 / AVC in units of blocks, it is possible to perform coding suitable for the properties of the image. Further, the present invention is particularly effective for encoding at a high bit rate, particularly lossless encoding.

  Embodiments of the present invention will be described below with reference to the drawings.

FIG. 3 conceptually shows the operation of the intra-screen predictive encoding process based on H.264 / AVC, which is a conventional technique. In H.264 / AVC, encoding is performed on an encoding target image according to the order of raster scanning (301), and predictive encoding is performed using a decoded image in an encoded image region.

Therefore, in the operation example (302) of the intra prediction encoding process, all the pixels on the same straight line with the prediction direction vector as the gradient are predicted from the same pixel. For example, as shown in (303) All the pixels B, C, D, and E of the conversion target block are subjected to predictive encoding with reference to the same pixel A, and a difference (prediction difference) from a value obtained by decoding the pixel A is calculated. In addition, the optimal prediction direction can be selected from several types of prediction direction candidates such as vertical, horizontal, and diagonal (block 304).
), The prediction difference and the value of the prediction direction are encoded.

On the other hand, in the decoding process, the reverse process of the above process may be performed using the decoded reference pixel and the prediction difference, and the decoded image is obtained by adding the prediction difference to the non-reference pixel along the prediction direction. (305). For example, in (306), all the pixels B, C, D, and E on the decoding target block are decoded by the sum of the decoded reference pixel A and the prediction difference. As above, H.264 / A
In the intra-screen predictive encoding process using VC, it is necessary to refer to pixels that are spatially separated in order to limit the referenced pixels to those that have already been encoded (for example, pixel E is predicted with reference to distant pixel A) Encoding has been performed.) However, there is a problem in that the correlation is lowered between distant pixels and the prediction accuracy is not sufficient.

FIG. 4 is a diagram conceptually showing the operation of the intra prediction encoding process according to the present invention.
In this case as well, the encoding target image is encoded according to the raster scan order (401).
However, in the present invention, in order to solve the above problem, it is possible to perform prediction by referring to spatially close pixels even if the encoding process is not completed. For example, in (402), prediction is always performed in the vertical direction with reference to adjacent pixels.

In this case, when referring to an unencoded area, the decoded pixel information cannot be used as in the prior art, and therefore prediction is performed using the value of the original image located at the same coordinates. For example, in (403), pixels C, D, and E are predicted using unencoded pixels B, C, and D, respectively. Difference between pixels in position (B-
A = b, CB = c, DC = d, ED = e). Furthermore, various prediction directions can be defined for the present invention (404), and the prediction difference and the value of the prediction direction are encoded.

When decoding, unlike the case of encoding, only the decoded image is used for the prediction process, and the decoded image is repeated by repeating the process of adding the prediction difference to the decoded image located at the immediately preceding coordinate along the prediction direction. Can be obtained (405). For example, in (406), pixels B, C, D, E are pixels A, B,
A decoded image can be acquired by adding the prediction difference (b ′, c ′, d ′, e ′) to the decoded image of C, D (A ′, B ′, C ′, D ′).

In the case of the present invention, if the quantization process is performed on the prediction difference at the time of encoding, a quantization error may be accumulated. For this reason, if the present invention is used, the prediction accuracy is improved and the prediction difference is reduced, but there may be a case where the quantization error increases. In such a case, encoding efficiency may be higher when predictive encoding is performed using the conventional technique.

Therefore, by using the predictive coding technique according to the present invention in combination with the conventional technique, a high compression rate can be realized according to the property of the image. For example, predictive encoding using the present technology and the conventional technology separately in units of blocks enables encoding suitable for the properties of the image. As the prior art, for example, it is effective to use intra-screen predictive encoding using H.264 / AVC shown in FIG. 3 or inter-screen predictive encoding using an image (decoded image) different from the encoding target image. .

FIG. 1 shows an embodiment of a moving picture coding apparatus according to the present invention. Similarly, the moving image coding apparatus includes an input image memory (102) that holds an input original image (101), and a motion search unit (103) that detects motion in units of blocks obtained by dividing the input image into small regions. In-screen prediction unit (104) that performs intra-screen prediction on a block basis, inter-screen prediction unit (105) that performs inter-screen prediction based on the detected amount of motion, and predictive coding means that matches the nature of the image A mode selection unit (106), a subtraction unit (107) for generating a prediction difference, a frequency conversion unit (108) that performs encoding on the prediction difference,
And a quantizing unit (109) and a variable length coding unit (11) for performing coding according to the probability of occurrence of a symbol
0), an inverse quantization processing unit (111) for decoding a prediction difference once encoded, and an inverse frequency transform unit (112), and for generating a decoded image using the decoded prediction difference Addition part (113)
And a reference image memory (114) for holding the decoded image and using it for later prediction.

The original image memory (102) holds one image from the original image (101) as an encoding target image, and divides it into fine blocks to divide the image into a motion search unit (103) and an in-screen prediction unit (104 ).

The motion search unit (103) calculates the motion amount of the corresponding block using the decoded image stored in the reference image memory (114), and passes it to the inter-screen prediction unit (105) as a motion vector.

The in-screen prediction unit (104) and the inter-screen prediction unit (105) execute the intra-screen prediction process and the inter-screen prediction process in blocks of several sizes, respectively, and the mode selection unit (106) Select predictive encoding means. Subsequently, the adder (107) generates a prediction difference by the optimum predictive encoding means and passes it to the frequency converter (108).

The intra-screen prediction unit (104) includes an improved intra-screen prediction encoding unit that performs predictive encoding processing with reference to a pixel that is always adjacent to an encoding target pixel that is a feature of the present invention, for example, H .264 / AVC
Includes a conventional intra-frame predictive encoding means.

The frequency conversion unit (108) and the quantization processing unit (109) each receive a DCT (Discrete) in units of blocks of a specified size for the transmitted difference image.
Frequency conversion such as Cosine Transformation (discrete cosine transformation) and quantization processing are performed, and the result is passed to the variable length coding processing unit (110) and the inverse quantization processing unit (111).

Further, in the variable length coding processing unit (110), the prediction difference information represented by the frequency transform coefficient is necessary for predictive decoding such as a prediction direction in intra prediction encoding and a motion vector in inter prediction encoding. Along with the additional information, variable length coding is performed based on the occurrence probability of symbols to generate a coded stream.

In addition, the inverse quantization processing unit (111) and the inverse frequency conversion unit (112) respectively perform inverse frequency conversion such as inverse quantization and IDCT (Inverse DCT) on the frequency conversion coefficient after quantization. To obtain the prediction difference and send it to the adder (113). Subsequently, the adder (113) generates a decoded image and stores it in the reference image memory (114).

FIG. 2 shows an embodiment of the moving picture decoding apparatus according to the present invention. For example, the moving picture decoding apparatus includes an encoded stream (201) generated by the moving picture encoding apparatus shown in FIG.
A variable length decoding unit (202) that follows the reverse procedure of variable length coding, an inverse quantization processing unit (203) for decoding a prediction difference, and an inverse frequency transform unit (204), Intra-screen prediction unit (205) and inter-screen prediction unit (206) for performing prediction, an addition unit (207) for acquiring a decoded image, and a reference for temporarily storing the decoded image An image memory (208) is included.

The intra-screen prediction unit (205) includes an intra-screen predictive decoding unit that always predicts a pixel to be decoded, which is a feature of the present invention with reference to adjacent pixels, and a conventional type based on, for example, H.264 / AVC. Intra-screen predictive encoding means is included.

The variable length decoding unit (202) decodes the encoded stream (201), and obtains additional information necessary for prediction processing, such as a frequency conversion coefficient component of a prediction difference and a prediction direction and a motion vector. The former prediction difference information is sent to the inverse quantization processing unit (203), and the latter additional information is sent to the intra prediction unit (205) or the inter prediction unit (206) depending on the prediction means. It is done. Subsequently, the inverse quantization processing unit (203) and the inverse frequency transform unit (204) perform inverse quantization and inverse frequency transform on the prediction difference information to perform decoding, while the intra prediction unit (205 ) Or the inter-screen prediction unit (206) executes a prediction process based on the additional information, and the addition unit (207) generates a decoded image and stores the decoded image in the reference image memory (208).

FIG. 5 shows a procedure for encoding one frame in the embodiment of the moving picture encoding apparatus shown in FIG. First, the following processing is performed for all blocks existing in the frame to be encoded (501). In other words, the prediction coding process is performed once for all coding modes (combination of prediction method and block size) for the corresponding block, the prediction difference is calculated, and the one with the highest coding efficiency is selected from among them. To do. As a prediction processing method, in addition to the method according to the present invention (hereinafter referred to as “the improved intra prediction encoding process” (505)), for example, H.264 / A
Execute the intra-screen prediction method (hereinafter referred to as “conventional intra-screen predictive encoding process” (506)) and inter-screen predictive encoding process (507) used in VC, and select the optimal one from them. Thus, encoding can be performed efficiently according to the nature of the image. Furthermore, when selecting the one with the highest encoding efficiency from among the above-mentioned many encoding modes (508), for example, an RD-Optimization method for determining the optimal encoding mode from the relationship between image quality distortion and code amount is used. By using this, encoding can be performed efficiently. Details of the RD-Optimization method are described in Non-Patent Document 1.

Subsequently, frequency conversion (509) and quantization processing (510) are performed on the prediction difference generated in the selected encoding mode, and further, variable length encoding is performed to generate an encoded stream (511). ). On the other hand, the quantized frequency transform coefficient is subjected to inverse quantization processing (512) and inverse frequency transform processing (513) to decode the prediction difference, generate a decoded image, and store it in the reference image memory ( 514). When the above processing is completed for all the blocks, the encoding for one frame of the image is completed (515).

FIG. 6 shows the detailed processing procedure of the improved intra prediction encoding process (505). Here, (601) for all predefined prediction directions, such as vertical, horizontal, diagonal,
The following processing is performed in the order along the prediction direction for all pixels present in the block (6
02). That is, if the corresponding pixel belongs to the block boundary (603), the difference between the decoded image of the previous coordinate and the original image of the current coordinate is calculated (604), and if not, the original image of the previous coordinate and the current coordinate The difference between the original images is calculated (605). If the above process is completed for all prediction directions, a prediction difference for one block can be acquired, and the prediction encoding process is terminated (60
6).

FIG. 7 shows a procedure for decoding one frame in the embodiment of the moving picture decoding apparatus shown in FIG. First, the following processing is performed for all blocks in one frame (
701). That is, variable length decoding processing is performed on the input stream (702), and inverse quantization processing (
703) and inverse frequency transform processing (704) are performed to decode the prediction difference. Subsequently, an improved intra prediction decoding process (707) depending on which method is used for predictive encoding of the target block
Then, conventional intra-screen predictive decoding processing (708) and inter-screen predictive decoding processing (709) are performed, and a decoded image is acquired and stored in the reference image memory. If the above process is completed for all blocks,
Decoding of one image frame is completed (710).

FIG. 8 shows a detailed processing procedure of the improved intra prediction decoding process (707). Here, the following processing is performed on all the pixels present in the block in the order along the prediction direction (801). That is, the sum of the decoded image of the previous coordinate and the prediction difference of the current coordinate is calculated (
802). If the above processing is completed for all the pixels, a decoded image for one block can be acquired, and the predictive decoding processing ends (803).

In the improved intra prediction encoding process according to the above embodiment, a decoded image of an encoded pixel is used for prediction of a pixel located at a block boundary. This is to prevent the quantization error in the target block from propagating to another block. However, in this case, there is a possibility that the prediction error becomes large at the block boundary. Therefore, in order to reduce the prediction error, prediction encoding may be performed on the pixels located at the block boundary using the original image of the immediately preceding coordinates.

  Further, in the improved intra prediction encoding process according to the present embodiment, the value of the reference pixel at the time of encoding is different from that at the time of decoding, so that a quantization error is likely to be accumulated. Therefore, in order to reduce this quantization error, as another configuration example of the improved intra prediction encoding process, the pixel value of the reference pixel may be smoothed when performing prediction. In this case, the error is diffused and the prediction accuracy is increased. This effect is particularly high at low bit rates. Hereinafter, another example of the configuration will be described.

  FIG. 12 is a configuration example when the above smoothing process is used in the improved intra prediction encoding process shown in FIG. In the configuration example of FIG. 12, the pixel value of the reference pixel is smoothed. Here, for all predefined prediction directions such as vertical, horizontal, and diagonal (1201), the following processing is performed in the order along the prediction direction for all pixels present in the block (1202). . That is, if the corresponding pixel belongs to the block boundary (1203), the decoded image at the immediately preceding coordinates is smoothed (1204), and then the difference between the decoded image after smoothing and the original image at the current coordinates is calculated (1206). On the contrary, if it does not belong to the block boundary, the original image at the immediately preceding coordinates is smoothed (1205), and the difference between the original image after smoothing and the original image at the current coordinates is calculated (1207). If the above processing is completed for all prediction directions, a prediction difference for one block can be acquired, and the prediction encoding processing is terminated (1208).

  Here, the smoothing process in step 1204 and step 1205 in FIG. 12 will be described with reference to FIG. FIG. 11 shows an embodiment of the smoothing process of the reference pixel.

  The smoothing process in step 1204 of FIG. 12 is shown as a smoothing process for the pixel F ′ in FIG. 11, for example. That is, if it is determined in step 1203 of FIG. 12 that the pixel to be predicted is a pixel belonging to the block boundary (pixel J in this example), the decoded image of the pixel at the immediately previous coordinate (pixel F ′) and its surrounding pixels Is smoothed. In this example, the pixel value S (F ′) after smoothing of the pixel F ′ to be smoothed is the pixel value of the decoded image of the pixel F ′ and each pixel (G ', H') is calculated as a weighted average with the pixel value of the decoded image. Here, S (X) is an example of a function for calculating a pixel value obtained by smoothing pixel values of a decoded image of a plurality of pixels for the pixel X.

  The smoothing process in step 1205 of FIG. 12 is shown as a smoothing process for the pixel J in FIG. That is, if it is determined in step 1203 of FIG. 12 that the prediction target pixel is a pixel that does not belong to the block boundary (in this example, pixel L), the original image of the pixel at the immediately previous coordinate (pixel J) and the surrounding pixels Is smoothed. In this example, the pixel value T (J) after smoothing of the pixel J to be smoothed is the pixel value of the original image of the pixel J and each pixel (I, K) located on both sides of the pixel J. It is calculated as a weighted average with the pixel value of the original image. Here, T (X) is an example of a function for calculating a pixel value obtained by smoothing pixel values of an original image of a plurality of pixels for the pixel X.

  In the above two smoothing processing examples, an example of a weighted average is used. However, the value of the function whose variable is the pixel value (the decoded image or the original image) of multiple pixels located around the pixel to be smoothed (which may be adjacent or distant) Any conventional smoothing method may be used as long as the method is used to calculate the pixel to be smoothed.

  According to the image encoding device and the image encoding method that perform the improved intra prediction encoding process using the smoothing process described above, it is possible to reduce the quantization error at the time of decoding, and to reduce the amount of codes Thus, it is possible to generate an encoded stream that can decode a higher quality video.

  Next, a configuration example in the case where the smoothing process is used in the improved intra prediction decoding process shown in FIG. 8 will be described with reference to FIG. Here, the following processing is performed on all the pixels present in the block in the order along the prediction direction (1301). That is, the decoded image at the immediately preceding coordinates is smoothed (1302), and the sum of the smoothed decoded image and the prediction difference between the current coordinates is calculated (1303). If the above process is completed for all the pixels, a decoded image for one block can be acquired, and the predictive decoding process ends (1304). The smoothing process shown in Step 1302 of FIG. 13 may be performed, for example, in the same way as when the pixel value S (F ′) is calculated in the smoothing process of the pixel F ′ in Step 1204 of FIG.

  According to the image decoding apparatus and the image decoding method that perform the improved in-screen predictive decoding process using the smoothing process described above, it becomes possible to further reduce the quantization error and to display a higher quality video. Decoding is possible.

In the embodiment, DCT is cited as an example of frequency conversion, but DST (Discrete Sine T
ransformation (Discrete sine transformation), WT (Wavelet Transformation)
, DFT (Discrete Fourier Transformation), KLT (Karhunen-Loeve
Any transformation may be used as long as the orthogonal transformation is used to remove the correlation between pixels, such as Transformation (Karunen-Reeve transformation), and the prediction difference itself may be encoded without performing frequency transformation. Furthermore, variable length coding is not particularly required.

FIG. 9 shows a configuration example of a part related to encoding parameters to be set in units of blocks in an encoded stream generated when the present invention is used. Here H.
Similar to the processing unit in H.264 / AVC, a case has been described in which the encoding mode is determined in units of fixed-length macroblocks. The macroblock can be divided into finer blocks, and predictive coding is performed in units of the divided blocks. At this time, the macroblock number (901) for specifying the coordinates for each macroblock, the encoding mode number (902) indicating the prediction method and the block size, and the motion vector and intra-screen for the inter-screen prediction, for example Additional information (903) necessary for prediction, such as a prediction direction in prediction, and prediction difference information (904) are variable-length-coded and stored. In particular, the encoding mode number (902) may be assigned sequentially to all prediction means, or may be represented by different bits for each prediction means.

FIG. 10 shows another configuration example for the part related to the encoding parameter to be set in units of blocks. In this example, the macroblock number (1001) and the prediction method (1002)
On the other hand, a flag (1003) indicating whether or not to perform prediction using adjacent pixels according to the present invention is added, and additional information (1004) and prediction difference information (1005) are stored.

Although the present embodiment describes the case of encoding a moving image, the present invention is also effective for encoding a still image. That is, if the motion search unit (103) and the inter-screen prediction unit (105) are excluded from the block diagram of FIG. 1, this corresponds to a block diagram of an encoding device specialized for still images.

FIG. 1 is a block diagram of an image encoding apparatus used in this embodiment. FIG. 2 is a block diagram of an image decoding apparatus used in this embodiment. FIG. 3 is a conceptual explanatory diagram of intra prediction encoding processing used in H.264 / AVC. FIG. 4 is a conceptual explanatory diagram of the intra prediction encoding process used in the present embodiment. FIG. 5 is a flowchart of the image encoding apparatus used in this embodiment. FIG. 6 is a detailed flowchart of the image encoding apparatus used in this embodiment. FIG. 7 is a flowchart of the image decoding apparatus used in this embodiment. FIG. 8 is a detailed flowchart of the image decoding apparatus used in this embodiment. FIG. 9 is a configuration example of an encoded stream used in the present embodiment. FIG. 10 is a configuration example of an encoded stream used in this embodiment. FIG. 11 is a conceptual explanatory diagram of the smoothing process used in the present embodiment. FIG. 12 is a detailed flowchart of the image encoding apparatus used in this embodiment. FIG. 13 is a detailed flowchart of the image decoding apparatus used in this embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Input image 102 Current image memory 103 Motion search part 104 Intra-screen prediction part 105 Inter-screen prediction part 106 Mode selection part 107 Subtraction part 108 Frequency conversion part 109 Quantization process part 110 Variable length encoding part 111 Inverse quantization process part 112 Inverse frequency conversion unit 113 Addition unit 114 Reference image memory 201 Encoded stream 202 Variable length decoding unit 203 Inverse quantization processing unit 204 Inverse frequency conversion unit 205 Intra-screen prediction unit 206 Inter-screen prediction unit 207 Addition unit 208 Reference image memory 301 306 Explanation diagrams of intra prediction encoding processing by H.264 / AVC 401 to 406 Explanation diagrams of intra prediction encoding processing according to the present invention 501 to 515 Flowchart blocks 601 to 606 Flowchart blocks 701 to 710 Flowchart blocks 801-803 Block chart block 901-904 Information stored in stream 1001-1005 Information stored in stream 1201-1208 Flowchart block 1301-1304 Flowchart block

Claims (16)

In order to perform encoding according to the probability of occurrence of a symbol, an intra-screen prediction unit that performs intra-screen prediction in block units and calculates a prediction difference, a frequency conversion unit and a quantization unit that perform encoding on the prediction difference, and In the video encoding device having the variable length encoding unit,
The moving picture encoding apparatus, wherein the intra prediction unit includes an intra prediction encoding unit that performs predictive encoding processing with reference to a pixel that is always adjacent to a pixel to be encoded. The moving image encoding device according to claim 1,
In the intra prediction unit, a plurality of prediction directions are defined, and a difference between adjacent pixels along a corresponding prediction direction is calculated at the time of predictive encoding. In the moving image encoding device according to claim 1 or 2,
In the in-screen prediction unit, in addition to the in-screen predictive encoding means for performing predictive encoding processing with reference to pixels that are always adjacent to pixels to be encoded, in-screen predictive processing is executed. A predictive encoding unit, and an encoding mode selection unit that selects an optimal result from a plurality of intra-screen prediction encoding units of the intra-screen prediction unit, in units of blocks or images. A moving image encoding apparatus comprising: Performs intra-frame prediction on a block basis, calculates the prediction difference, performs frequency conversion and quantization to encode the prediction difference, and performs variable length coding to perform coding according to the probability of occurrence of the symbol In the moving image encoding method to be performed,
An intra-frame prediction encoding method for performing intra-prediction encoding processing for performing predictive encoding processing with reference to a pixel that is always adjacent to a pixel to be encoded in the intra-frame prediction. In the moving image encoding method according to claim 4,
A plurality of prediction directions are defined at the time of intra prediction, and a difference between adjacent pixels along the corresponding prediction direction is calculated at the time of prediction encoding. Method. In the moving image encoding method according to claim 4 or 5,
In the case of the intra prediction, in addition to the intra prediction encoding for executing the predictive encoding process with reference to the pixels that are always adjacent to the pixel to be encoded, a screen for executing a different predictive encoding process A moving picture code characterized by performing intra prediction encoding and selecting an optimum one from a plurality of results encoded by intra prediction encoding at the time of the predictive encoding in units of blocks or images Method. A variable-length decoding unit that performs the reverse procedure of variable-length coding, an inverse quantization processing unit and an inverse frequency transform unit for decoding a prediction difference, and intra-frame prediction for each block to obtain a decoded image In a video decoding device having an intra-screen prediction unit for
An intra-screen prediction unit includes an intra-screen prediction decoding unit that always predicts a pixel to be decoded with reference to adjacent pixels. The moving picture decoding apparatus according to claim 7,
A plurality of prediction directions are defined in the in-screen prediction unit, and a moving image is calculated by calculating a sum of a decoded image of a pixel adjacent to the corresponding prediction direction and a prediction difference in predictive decoding. Image decoding device. The moving picture decoding apparatus according to claim 7 or 8,
In addition to the intra-screen prediction decoding unit, the intra-screen prediction unit includes an intra-screen prediction decoding unit that executes different prediction decoding processes. In the plurality of intra-screen prediction decoding units, A moving picture decoding apparatus that decodes data that has been predictively encoded by a different method in units or images. Performs variable-length decoding following the reverse procedure of variable-length coding, performs inverse quantization processing and inverse frequency transform to decode the prediction difference, and obtains a decoded image by performing intra prediction in units of blocks In the video decoding method for performing intra-screen prediction for
An intra-picture prediction method for performing intra-prediction decoding that predicts a pixel to be decoded with reference to adjacent pixels at the time of intra-prediction. The moving picture decoding method according to claim 10, wherein
A plurality of prediction directions are defined when performing the intra-screen prediction, and at the time of predictive decoding, the prediction difference of the pixel to be decoded is added to the decoded image of the adjacent pixel along the corresponding prediction direction. A video decoding method characterized by the above. In the moving image decoding method according to claim 10 or 11,
In addition to intra-frame predictive decoding in which the pixel to be decoded is always predicted with reference to adjacent pixels, intra-frame predictive decoding in which predictive decoding processing is performed by different methods is possible. A moving picture decoding method, comprising decoding data that has been predictively encoded by a method different in units.   The moving picture encoding apparatus according to claim 1, wherein the intra-screen predictive encoding unit includes a pixel value of a pixel adjacent to the pixel to be encoded or a plurality of pixels located around the adjacent pixel. A moving picture coding apparatus characterized by referring to a smoothing function value having a variable of a pixel value of any one of a plurality of decoded pictures or a pixel value of an original picture.   5. The moving image encoding method according to claim 4, wherein in the intra prediction, a pixel value of a pixel adjacent to the pixel to be encoded or a plurality of pixels located around the adjacent pixel is calculated. A moving picture coding method characterized by referring to a smoothing function value having a variable of a pixel value of any one of a plurality of decoded pictures or a pixel value of an original picture.   The moving picture decoding apparatus according to claim 7, wherein the intra prediction decoding means includes a pixel value of a pixel adjacent to the pixel to be decoded or a plurality of pixels located around the adjacent pixel. A moving picture decoding apparatus characterized by referring to a smoothing function value having a pixel value of any one of a plurality of decoded pictures as a variable.   The moving picture decoding method according to claim 10, wherein in the intra prediction, a pixel value of a pixel adjacent to the pixel to be decoded or a plurality of pixels located around the adjacent pixel is selected. A moving picture decoding method characterized by referring to a value of a smoothing function whose variable is a pixel value of any one of a plurality of decoded pictures.

Images