JP2887177B2 - Method and apparatus for decoding moving image compression code - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for decoding a moving image compression code, and more particularly, to a method suitable for a case where an error occurs in a broadly defined transmission path such as a communication path or a storage medium.

[0002]

2. Description of the Related Art In digital communication, when transmitting data having an enormous amount of information such as a moving image,
Data is subjected to information compression to reduce redundancy.

Reference (1): "TTC standard (higher layer protocol coding method) JT-H.261 recommendation" FIG. 2 shows the configuration of a moving picture compression encoder / decoder specified in reference (1). Show. A video signal, which is the original data of a moving image, has a globally common intermediate format (CIF, QCI
F), and when this moving image data (one picker; one frame) is encoded, as shown in FIG. 3, a group of blocks (hereinafter referred to as GOB (grou
p of block), macro block (Macroblo
ck), and divided hierarchically into blocks of 8 × 8 pixels (Block).

When encoding is performed by the encoder shown in FIG. 2A, a video signal is input and the information is encoded by an information source encoder 201. The information source encoder 201 first performs motion compensation inter-frame prediction for each macroblock, and then quantizes the prediction error after performing orthogonal transform for each block. Prediction is usually performed between frames, but at the time of a scene change or the like, the original data of a moving image is orthogonally transformed as it is. The determination as to whether to orthogonally transform the motion-compensated inter-frame prediction error or to perform orthogonal transformation on the original data of the moving image as it is is determined by, for example, comparing the variance of the prediction error data and the original data. Discrete Cosine Transform (DCT) for orthogonal transform
Is used to convert from the representation in the pixel domain to the representation in the frequency domain. The data converted to the frequency domain representation and quantized is input to the video signal multiplexing encoder 202, subjected to variable length coding, and multiplexed with header information. The source coded and multiplexed data by the source coder 201 and the video signal multiplexing coder 202 are transmitted to the transmission buffer 20.
3, the transmission path encoder 204 performs transmission path encoding in the order shown in FIG. In order to control the transmission code amount, the state in the buffer is transmitted from the transmission buffer 203 to the encoding control 205, and the encoding control 205
Are the source coder 201 and the video signal multiplexing coder 2
02, a control signal is issued.

FIG. 4 shows a coded bit sequence and a coding syntax coded by a conventional moving picture coding method. The encoding syntax defines a variable-length encoding table, header information to be inserted, and a transmission order.

[0006] In FIG. 4, elements indicated by numerals are data types to be fixed-length coded, and other elements are data types to be variable-length coded using different variable-length coding tables. .

The frame start code PSC indicating the beginning of a frame is a special fixed-length code, and the same bit sequence as the frame start code PSC appears in encoded data other than the frame start code PSC. Not to be. The frame header information PHEAD includes information on the entire frame such as a frame number and picture type information, and is fixed-length coded.

[0008] The GOB start code GBSC indicating the beginning of the GOB is fixed-length coded. The GOB header information GHEAD includes information on one GOB, such as a GOB number indicating a position in a picture of the GOB and quantization characteristic information of the GOB, and is fixed-length encoded.

The macro block address MBA is GOB
Indicates the position of the macroblock in the macroblock, and the difference between the macroblock address and the previously coded macroblock address is variable-length coded. If there is no information (for example, the same as the previous frame), the macro block is not coded, and the macro block address MBA, the macro block type information MTYP
E, the macroblock quantization characteristics MQUANT, the motion vector information MVD, and the block information (TCOEFF, EOB) of the macroblock are not transmitted. The macroblock type information MTYPE indicates whether the prediction is between frames (hereinafter, referred to as an INTER mode), whether or not the original signal is orthogonally transformed (hereinafter, referred to as an INTRA mode), and is variable-length coded. The INTER mode is adopted, for example, when the motion is small (the difference between frames is small), and the INTRA mode is adopted when the motion is large. Macro block quantization characteristics MQUANT
Represents the quantization step size of the macroblock, and is fixed-length coded if different from the quantization step size of the previously encoded macroblock. The motion vector information MVD is obtained by subtracting the motion vector of the previous macroblock from the motion vector of the current macroblock, and is subjected to variable length coding. The significant block pattern CBP indicates the position of a block (significant block) to which at least one transform coefficient is transmitted, and is subjected to variable length coding.

The above-described transform coefficient TCOEFF is also subjected to variable length coding. The block end code EOB indicating the end of the block is fixed-length coded, but the transform coefficient TCOEFF
Is one of the codewords of the variable length code.

FIG. 5 shows a variable-length coding table of data types to be subjected to variable-length coding of macroblocks. Each data type is subjected to variable-length coding according to the variable-length table. Here, the transform coefficient TCOEFF is obtained by rearranging block data of 8 × 8 pixels quantized by orthogonal transformation in the order of lower frequency components in the horizontal and vertical directions, and the number of consecutive 0s (0 run) ) And a non-zero value (level) immediately after the variable-length coding.

FIG. 6 shows a multiplexing processing system diagram of the video signal multiplexing encoder 202 in the case of the conventional example.

In each of the frame layer, GOB layer, macroblock layer and block layer, multiplexing is performed in the direction of the arrow. Where the arrow branches, the mode (INT
RA, INTER, etc.), one of the arrows is selected. The looped part is the data (GO
B layer, macroblock layer, block layer, conversion coefficient TCO
Until EFF) ends, the arrow on the loop side is selected, and when that data ends, the non-loop arrow is selected. The GOB in the frame, the macroblock in the GOB, and the block in the macroblock are shown in FIG.
As shown in the figure, a GOB number, a macroblock address, and a block number are added in order from the upper left. GO
The B layer, the mask block layer, and the block layer are coded and multiplexed in ascending order of number (address).

When decoding is performed by the decoder shown in FIG. 2B, the transmission path is decoded by the transmission path decoder 209 in the order in which it is transmitted, and is transmitted to the video signal multiplexing decoder 207 through the reception buffer 208. input. The video signal multiplexing decoder 207 separates the header information and the like, decodes the variable-length code, and outputs the result to the information source decoder 206. The information source decoder 206 performs inverse quantization, inverse discrete cosine transform (inverse DCT), and motion compensation, and outputs a video signal.

Whether or not an error has occurred can be determined in the decoding of the variable length code by the video signal multiplexing decoder 207. For example, it can be assumed that an error is detected when a sequence that does not exist in the variable-length codeword of the variable-length coding table is input. When such an error occurs, the boundary of the variable-length code is lost and synchronization is lost.

If an error occurs, the macroblock address MBA is a value obtained by encoding a difference value from a previously encoded macroblock. From the point of occurrence, the next G
Until the OB start code GBSC or the frame start code PSC is detected, decoding cannot be performed. Therefore,
Data from the time when the error occurs to the GOB start code GBSC or the frame start code PSC is discarded. The part of the image data discarded due to the error is
Output the previous frame image as it is.

The operation of the conventional video signal multiplexing decoder 207 will be described with reference to the flowchart shown in FIG.

First, the header information is separated from the code sequence input to the video signal multiplexing decoder 207 and the variable length code is decoded (step B1). Thereafter, it is determined whether or not an error has occurred in decoding the variable length code (step B).
2). If an error has occurred, the next GOB start code GBSC or frame start code PSC is searched (step B3), and if the next GOB start code GBSC or frame start code PSC is detected (step B4), Data from the error occurrence point to the detected GOB start code GBSC or frame start code PSC is discarded (step B5). If an error occurs and data up to the next GOB start code GBSC or frame start code PSC is discarded, and if it is determined in step B2 that there is no error, a decoding result is output. (Step B
6) It is further determined whether or not the input has been completed (step B7). When the input has been completed, a series of processing is ended, and when the input is continued, the process returns to step B1 described above. Note that the next GOB start code G
The data up to the BSC or the frame start code PSC is discarded, but as the decoding result of the image data discarded due to the error, the data of the previous frame image is output as it is as described above (step B6).

[0019]

In the decoding method according to the moving picture coding syntax shown in FIG. 4 and the video signal multiplexing decoding operation shown in FIG. 7, when an error occurs,
When decoding the variable-length code, the boundary of the variable-length code word is lost, and synchronization is lost, so that decoding becomes impossible.
Therefore, information cannot be decoded until the next frame start code PSC or GOB start code GBSC is detected, and even if no error occurs in a part that cannot be decoded, data in that part must be discarded. There is a problem that can not be obtained.

As shown in FIG. 8, when an error occurs, generally, the decoding side cannot immediately detect that the error has occurred, and the decoded image data is output with the error. Therefore, this is a cause of image quality deterioration. For example, if an error is detected when a sequence that does not exist in the variable-length codeword of the variable-length coding table is input, the variable-length code at the time of the error even if an error has occurred earlier. If the word exists in the variable length coding table, the error cannot be detected, and the error is detected only after finding a variable length code word that does not exist in the variable length coding table.

[0021]

According to a first aspect of the present invention, there is provided an image area for transmitting information based on a difference between a current image and a previously transmitted reference image, and an image area for transmitting information only from the current image. In the moving picture compression code decoding method of decoding a moving picture compression code in which variable length codes are mixed with the above, when an error occurring in the transmission path is detected, the error is detected between the previous synchronization position and the next synchronization position. Discards image region information decoded only from information of the current image independent of the reference image, and generates a decoded image of the discarded image region and an image region that could not be decoded due to an error from the reference image It is characterized by the following.

According to a second aspect of the present invention, a variable length code for an image area for transmitting information based on a difference between a current image and a reference image transmitted before that and an image area for transmitting information only from the current image are used. In a moving picture compression code decoding method for decoding a moving picture compression code in which the self-synchronization recovery function of the variable-length code is available, when the error occurring in the transmission path is detected, The synchronization of the long code is recovered, and of the code data from the synchronization recovery to the next synchronization position,
An image area that transmits information based on a difference from a reference image, and is decoded only from information of a current image that is independent of the reference image during a period from a previous synchronization position until an error is detected. The information is discarded, and a decoded image of the discarded image area and an image area that cannot be decoded is generated from the reference image.

According to a third aspect of the present invention, a variable-length code for an image area for transmitting information based on a difference between a current image and a reference image transmitted earlier and an image area for transmitting information only from the current image are used. In a moving picture compression code decoding device for decoding mixed moving picture compression codes, an error detection means for detecting an error occurring in a transmission path, and when an error is detected, from an earlier synchronization position to a next synchronization position. A discarding unit for discarding image region information decoded only from information of the current image independent of the reference image, and generating a decoded image of the discarded image region and the image region that could not be decoded from the reference image. Image interpolating means for performing the operation.

According to a fourth aspect of the present invention, a variable-length code for an image area for transmitting information based on a difference between a current image and a reference image transmitted before that and an image area for transmitting information only from the current image are used. In a moving picture compression code decoding apparatus for decoding a moving picture compression code in which the self-synchronization recovery function of the variable length code is available, error detection means for detecting an error occurring in a transmission path, Synchronization recovery means for recovering the synchronization of the lost variable-length code when detected,
A reference image using area decoding unit that decodes an image area transmitting information based on a difference from a reference image among code data up to the next synchronization position after synchronization recovery, and detects an error from a previous synchronization position. In the meantime, the discarding means for discarding the image area information decoded only from the information of the current image independent of the reference image, and the decoded image of the discarded image area and the image area that could not be decoded from the reference image. And an image interpolating means for generating.

According to the first to fourth aspects of the present invention, when the information of the image area is incorrect, the image quality of the decoded image is significantly deteriorated, and the image area information decoded only from information independent of the reference image is discarded. However, since the decoded image of the discarded image region and the image region that could not be decoded due to the error are generated from the reference image, even if the received moving image compression code has an error, the image quality of the decoded image is extremely deteriorated. Can be held down slightly.

[0026]

BEST MODE FOR CARRYING OUT THE INVENTION

(A) First Embodiment Hereinafter, a first embodiment of a method and apparatus for decoding a moving image compression code according to the present invention will be described in detail with reference to the drawings.

FIG. 9 is a block diagram showing the configuration around the information source decoder (corresponding to the information source decoder 206 in FIG. 2) of the moving picture compression code decoder according to the first embodiment.

Referring to FIG. 9, an information source decoder according to the first embodiment includes an information source decoder main body 11 for decoding a video signal from a sequence input from a video signal multiplexing decoder (see FIG. 2). A reference frame image memory 12 for storing a decoded image of a reference frame (for example, a previous frame) used in a decoding process of the information source decoder body 11, and a decoding operation of a frame being decoded. And an INTRA for storing which mode of INTRA / INTER each macroblock was decoded at the time of decoding the current frame.
/ INTER mode memory 14.

When the reference frame is the previous frame, the reference frame image memory 12 and the current frame image memory 13 correspond to, for example, two frame image memories, and store the current frame image during decoding. The frame image memory becomes the current frame image memory 13, the other frame image memory becomes the reference frame image memory 12, and the frame image memory that becomes the reference frame image memory 12 and the current frame image memory 13 changes for each frame. However, functionally, the reference frame image memory 12 and the current frame image memory 13 exist in any case.

Next, the decoding method of the first embodiment, in other words, the operation of the information source decoder main body 11 will be described.

The sequence input from the video signal multiplexing decoder is sequentially decoded into a macroblock image in the information source decoder body 11 while referring to the reference frame image memory 12 for each GOB, and the current frame image memory 13 is stored. At this time, whether each macroblock is decoded in the INTRA mode or in the INTER mode is determined by I
It is stored in the NTRA / INTER mode memory 14.

When the video signal multiplexing decoder or the information source decoder main body 11 detects that an error has occurred in a certain GOB, the next synchronization word (for example, the frame start code P
The input data up to the SC or GOB start code GBSC) is discarded, and the macroblock image stored in the INTRA / INTER mode memory 14 as decoded in the INTRA mode among the macroblocks that could be decoded until an error is detected is also discarded. I do. After that, an image of a macroblock for which decoding processing is not executed and an image of a macroblock discarded due to the INTRA mode are generated from the reference frame images stored in the reference frame image memory 12.

The INTRA macroblock is discarded only in the GOB in which an error has been detected.

FIG. 1 is a flowchart showing the flow of processing for each GOB among the processing of the decoding method of the first embodiment.

First, each macro block (MB) is decoded sequentially from the beginning of the GOB (step S11). And
It is determined whether an error has been detected (step S1).
2).

If no error is detected in each macroblock, the decoded macroblock image is stored in the current frame image memory 13 (step S13), and mode information indicating whether the image was decoded in the INTRA mode or in the INTER mode at this time. Is stored in the INTRA / INTER mode memory 14 (step S14).

The loop processing of steps S11 to S14 is repeated while confirming the end of the GOB until the end of the GOB (step S15). GOB in this way
Is detected, there is no error in the current GOB, decoding is completed, all decoded macroblock images are stored in the current frame image memory 13, and the process proceeds to the next GOB. Will be.

If an error is detected in step S12 before the end of the GOB is detected, the next synchronizing word (eg, frame start code PSC or GOB start code GBSC)
The input information up to this point is discarded (step S16), and among the macroblocks that can be decoded, the macroblock stored in the INTRA / INTER mode memory 14 as being in the INTRA mode is discarded (step S17).

Then, the image of the macroblock from which the decoded image has been discarded and the image of the macroblock for which the decoding process has not been executed because the input data is incorrect are extracted from the reference frame image stored in the reference frame image memory 12. It is generated (for example, taken out as it is) and stored in the current frame image memory 13 (step S18), and the processing of the GOB in which the error is detected is ended.

It should be noted that even if an image is obtained by executing a decoding process as an image after the macroblock in which an error is detected, the image quality deteriorates very much. Of course, it is preferable from the viewpoint that a strange image area does not exist (complete) as data, and as described above, it is conventionally employed.

Therefore, in the following, only the macroblock image decoded in the INTRA mode among the macroblocks before the macroblock in which the error is detected is discarded and the interpolation is performed using the reference frame image. Will be described.

As shown in FIG. 8 described above, when an error occurs, generally, the decoding side cannot immediately detect that the error has occurred, and the detection position is later than the occurrence position.
Between the occurrence position and the detection position, a decoded macroblock image is output with an error.

For this reason, it is preferable to interpolate a reference frame image as a decoded image between the occurrence position and the detection position, but since the occurrence position cannot be specified in the decoder, only the interpolation between the occurrence position and the detection position is performed. It is not possible. In addition, using the reference frame image as all the macroblock images before the occurrence position means that some macroblock images can be decoded normally, and such a normal macroblock image generally has the occurrence position and Since the area is larger than the erroneous macroblock image corresponding to between the detection positions, the image quality as a whole is rather deteriorated, which is not preferable.

Therefore, there is a demand for a method capable of improving the image quality as a whole by treating all the input data of the part before the error detection position (the normal part and the part between the error occurrence position and the detection position) in the same manner.

By the way, when the erroneously decoded macro block image is decoded in the INTRA mode,
When decoding is performed in the NTER mode, image quality of the former decoded image is significantly deteriorated.

An erroneously decoded macroblock image is I
When decoded as the NTER mode, the INTER mode is a mode for transmitting the difference from the reference image.
A decoded image obtained by adding noise to the reference frame image (for example, the previous frame) is obtained. On the other hand, when the erroneously decoded macroblock image is decoded in the INTRA mode, the erroneous INTRA macroblock image is decoded in the INTRA mode, irrespective of the reference frame image, from the input data of the current frame image itself. It is very likely that an image will be decoded into an image that is completely different from the surrounding macroblocks.

FIG. 10 schematically shows this state. As shown in FIG. 10 (b), when an error occurs, a macroblock that is erroneously decoded is INTRA
When the mode is decoded, the image quality is extremely deteriorated.

When the above points are arranged, the following can be understood from the aspect of image quality. The region decoded in the INTER mode is such that even if it is erroneous, noise is superimposed on the reference frame image, and in consideration of the image quality deterioration due to interpolation of the normal region using the reference frame image, an error occurs. In a case where a part before the detection position (a normal part and a part between the error occurrence position and the detection position) is similarly treated, it is preferable that the region decoded in the INTER mode is not interpolated using the reference frame image. .
On the other hand, when the region decoded in the INTRA mode is incorrect, the image quality deteriorates significantly when the region is erroneous, and the image quality deterioration due to interpolation of the normal region using the reference frame image is caused by the correlation between the successive frame images. From the error detection position (normal part,
In addition, when treating the portion between the error occurrence position and the detection position in the same manner, it is preferable that the macroblock decoded in the INTRA mode is interpolated (replaced) using the reference frame image.

Therefore, in the first embodiment, among the macroblocks before the macroblock in which the error is detected, the macroblock image decoded in the INTRA mode is discarded, and interpolation is performed using the reference frame image. It was decided to.

As described above, according to the first embodiment,
Not only the macroblock after the error is detected, but also before the error is detected, the image quality is severely degraded when an error is detected.
Since the decoded image of the NTRA macroblock is discarded and the image of the discarded macroblock is interpolated using the reference frame image, even if an error occurs, the overall image quality degradation can be kept smaller than before. it can.

(B) Second Embodiment Next, a second embodiment of a moving picture compression code decoding method and decoding apparatus according to the present invention will be described in detail with reference to the drawings.

Conventionally, as a moving image encoding / decoding method, a variable length code having a self-synchronization recovery function is used for coding, and when an error is detected during decoding, a variable-length code having a self-synchronization recovery function is searched for. In this way, even if an error occurs and the synchronization of the variable length code is lost, the synchronization is restored before the next synchronization word is reached while decoding continues, and the decoded data after the synchronization recovery can be used. Has already been proposed by the same applicant (Japanese Patent Application No. 6-247976 and the drawings).

The decoding method and the decoding apparatus according to the second embodiment relate to decoding of a moving picture compression code using a variable length code having such a self-synchronization recovery function.

The conventional decoding method of a moving picture compression code using a variable-length code having a self-synchronization recovery function has an effect of reducing the amount of data that cannot be used for decoding by self-synchronization recovery not depending on a synchronization word. However, there is a problem that the amount of suppression of image quality deterioration due to errors is still low.

As shown in FIG. 11A, a measure was taken to discard the data from the error detection position to the self-synchronization recovery and replace it with the reference frame image. Is present, and it cannot be avoided that image quality deteriorates. That is, FIG.
As shown in (b), "Problems to be solved by the invention"
In the same manner as described in the section, an error part may exist before the error detection position. In addition, depending on a data change due to an error, a variable length code having a self-synchronization recovery function may accidentally occur in an error portion. In such a case,
The self-synchronization recovery is erroneously determined, and erroneous data thereafter is used for decoding.

The information source decoder of the moving picture compression code decoder according to the second embodiment has the same configuration as the configuration shown in FIG. 1 according to the first embodiment. That is,
Information source decoder body 11, reference frame image memory 12,
Current frame image memory 13 and INTRA / INTER
A mode memory 14 is provided.

Note that the information source decoder and the preceding video signal multiplexing decoder correspond to the self-synchronization recovery function, which is different from the first embodiment.

Next, the decoding method of the second embodiment, in other words, the operation of the information source decoder main body 11 will be described.

The sequence input from the video signal multiplexing decoder (see FIG. 2) is read by the information source decoder body 11 while referring to the reference frame image memory 12 for each GOB.
They are sequentially decoded into macroblock images and stored in the current frame image memory 13. At this time, each macro block is I
Whether the data is decoded in the NTRA mode or the INTER mode is stored in the INTRA / INTER mode memory 14.

When the video signal multiplexing decoder or the information source decoder main body 11 detects that an error has occurred in a certain GOB, it performs synchronization recovery processing and discards input data until the synchronization recovery. After the synchronization recovery, among the decoded data up to the next synchronization word (for example, the frame start code PSC or the GOB start code GBSC), the macroblock in the INTER mode is decoded into a macroblock image and stored in the current frame image memory 13. . In other words, the decoded image of the macroblock related to the INTRA mode has been discarded between the synchronization recovery and the next synchronization word.
Also, among the macroblocks that could be decoded until an error was detected, the INTRA / INTER mode memory 14
The macroblock image stored as decoded in the TRA mode is discarded. Thereafter, an image of a macroblock in which decoding processing has not been performed due to an error in the input data and a macroblock in which the decoded image has been discarded due to the INTRA mode are generated from the reference frame image stored in the reference frame image memory 12. (Interpolate).

The INTRA macroblock is discarded only in the GOB where an error has been detected.

FIG. 12 is a flowchart showing the flow of processing for each GOB in the processing of the decoding method according to the second embodiment.

First, each macro block (MB) is decoded sequentially from the beginning of the GOB (step S21). And
It is determined whether an error has been detected (step S2).
2). If no error is detected in each macroblock, the decoded macroblock image is stored in the current frame image memory 13.
(Step S23), and at this time INT
The mode information indicating whether the decoding is performed in the RA mode or the INTER mode is stored in the INTRA / INTER mode memory 14 (step S24).

The loop processing consisting of the above steps S21 to S24 is repeated while confirming the end of the GOB until the end of the GOB (step S25). GOB in this way
Is detected, there is no error in the current GOB, decoding is completed, all decoded macroblock images are stored in the current frame image memory 13, and the process proceeds to the next GOB. Will be. As described above, the processing when no error occurs is the same as in the first embodiment.

If an error is detected in step S22 before the end of the GOB is detected, first, a synchronization recovery process is executed (step S26). The method described in Japanese Patent Application No. 6-247976 and the drawings can be applied to the synchronization recovery processing.

Next, after synchronization recovery, the next synchronization word (for example, frame start code PSC or GOB start code GBSC)
Of the input data up to this point, the data relating to each macroblock in the INTER mode is decoded and stored in the current frame image memory 13 (step S27). If the synchronization has not been restored and the next synchronization word has been reached, it is assumed that there is no data up to the next synchronization word after the synchronization recovery, and no decoding process is executed in step S27. Next, among the macroblocks decoded by the loop processing of steps S21 to S25, the macroblock image recorded in the INTRA / INTER mode memory 14 as being in the INTRA mode is discarded (step S28). Finally, an image of the macroblock from which the decoded image has been discarded and the remaining macroblocks that have not been decoded due to a data error are generated from the data stored in the reference frame image memory 12 and stored in the current frame image memory 13. (Step S29), the process of the GOB in which the error is detected ends.

In the second embodiment, only the image of the macro block in the INTER mode is decoded after synchronization recovery, and interpolation is performed using the reference frame image as the image of the macro block in the INTRA mode. This is the same as the reason for performing the same processing before the error detection position described in the first embodiment.

As described above, according to the second embodiment,
Since an image of an erroneous INTRA macroblock whose image quality is significantly deteriorated is discarded, there is no extreme image quality deterioration even when an error occurs. In addition, since the INTER macroblocks other than the INTRA macroblocks, which are highly likely to be degraded in image quality among the codes after synchronization recovery that are highly likely to be correctly decoded, are used for the decoded image, the image quality can be improved as compared with the related art.

(C) Other Embodiments Each of the above embodiments decodes a moving image compression code according to the TTC standard (higher layer protocol coding method) JT-H.261 recommendation or a partially improved moving image compression code. However, it goes without saying that the present invention can be applied to decoding of other moving image compression codes. For example, the number of layers for dividing the frame image may be smaller than that described above. Further, a method other than the discrete cosine transform may be applied as the orthogonal transform method. Further, the reference image may be the current image (line) above the encoding position. However, it is necessary that an area for encoding the difference from the reference image and an area for encoding without using the reference image coexist.

In the above description, an error is detected by finding a non-existent variable-length codeword.
The error detection method is not limited to this. For example, when decoding a variable length code (TCOEFF) of a transform coefficient portion, if more data than a predetermined number is obtained, an error may be detected.

[0071]

As described above, according to the present invention, when the information of the image area is incorrect, the image quality of the decoded image is significantly deteriorated, and the image area information decoded only from the information independent of the reference image. Is discarded, and the decoded image of the discarded image region and the decoded image of the image region that could not be decoded due to an error are generated from the reference image. Can be suppressed very slightly.

[Brief description of the drawings]

FIG. 1 is a flowchart illustrating a decoding process according to a first embodiment.

FIG. 2 is a block diagram illustrating a configuration of an encoder / decoder of a moving image compression code.

FIG. 3 is an explanatory diagram of hierarchical division of moving image data.

FIG. 4 is an explanatory diagram illustrating an example of a bit sequence configuration of a moving image compression code.

FIG. 5 is an explanatory diagram showing an example of a variable-length encoding table for each data type of the moving image compression code shown in FIG. 4;

FIG. 6 is an explanatory diagram showing a multiplexing processing system of the moving image compression code shown in FIG. 4;

FIG. 7 is a flowchart showing a conventional decoding process.

FIG. 8 is an explanatory diagram of a conventional problem.

FIG. 9 is a block diagram illustrating a configuration of an information source decoder according to the first embodiment.

FIG. 10 is an explanatory diagram illustrating a difference in decoded image quality between an INTER mode and an INTRA mode when an error occurs.

FIG. 11 is an explanatory diagram of a problem of a conventional decoding method premised on the second embodiment.

FIG. 12 is a flowchart illustrating a decoding process according to the second embodiment.

[Explanation of symbols]

11: information source decoder main body, 12: reference frame image memory, 13: current frame image memory, 14: INTRA /
INTER mode memory.

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Toshihisa Nakai 1-7-12 Toranomon, Minato-ku, Tokyo Oki Electric Industry Co., Ltd. (58) Field surveyed (Int. Cl. ⁶ , DB name) H04N 7 / 24-7/68

Claims (4)

(57) [Claims] 1. A moving image compression system in which variable-length codes are mixed in an image area for transmitting information based on a difference between a current image and a reference image transmitted before that, and an image area for transmitting information only from the current image. In the decoding method of a moving image compression code for decoding a code, when an error occurring in a transmission path is detected, information of a current image independent of a reference image from a previous synchronization position to a next synchronization position. Decoding of a video compression code, wherein a decoded image of a discarded image region and an image region that could not be decoded due to an error is generated from a reference image Method. 2. An image area for transmitting information based on a difference between a current image and a reference image transmitted before that, and an image area for transmitting information only from the current image, in which variable-length codes are mixed. In a moving picture compression code decoding method for decoding a moving picture compression code in which a self-synchronization recovery function of a variable length code is available, when an error occurring in a transmission path is detected, synchronization of the lost variable length code is performed. After the synchronization recovery, in the code data up to the next synchronization position, the image area transmitting the information based on the difference from the reference image is decoded, and until the error is detected from the previous synchronization position. Discarding image region information decoded only from information of a current image independent of a reference image, and generating a decoded image of the discarded image region and an image region that could not be decoded from the reference image. Video compression mark Method of decoding. 3. A moving picture compression system in which variable-length codes are mixed in an image area for transmitting information based on a difference between a current image and a reference image transmitted before that, and an image area for transmitting information based on only the current image. In a moving picture compression code decoding apparatus for decoding a code, an error detection means for detecting an error occurring in a transmission path, and when an error is detected, a reference is provided between a previous synchronization position and a next synchronization position. Discarding means for discarding image area information decoded only from information of the current image independent of the image; and image interpolating means for generating a decoded image of the discarded image area and the image area that could not be decoded from the reference image. An apparatus for decoding a moving image compression code, comprising: 4. An image area for transmitting information based on a difference between a current image and a reference image transmitted before that, and an image area for transmitting information only from the current image, and variable-length codes are mixed. In a moving picture compression code decoding apparatus for decoding a moving picture compression code capable of utilizing a self-synchronization recovery function of a variable length code, an error detection means for detecting an error occurring on a transmission line, and Synchronization recovery means for recovering the synchronization of the lost variable-length code, and a reference image for decoding an area transmitting information based on a difference from the reference image in code data up to the next synchronization position after the synchronization recovery. Use area decoding means; discard means for discarding an image area decoded only from information of a current image independent of a reference image until an error is detected from a previous synchronization position; and discarded image area and decoding Can Bought and the decoded image of the image area, the image interpolation means and decoding apparatus of the moving image compression encoding, characterized by having a generated from the reference image.