JPH0884334A - Decoding method for image signal and its device - Google Patents

Abstract

PURPOSE: To correct data with less deterioration in image quality by selecting correction data visually unremarkable due to replacement correction depending on the occurrence state with respect to code error data whose error correction is disable. CONSTITUTION: A coded data correction section 7 is provided between a multiplex video decoding section 3 and a video decoding section 4 to apply correct processing to code error data in which error correction of a video coding signal S3 is disable. In this case, a code error form is classified into a burst error and a random error depending on the occurrence state of code error data in which error correction of the video coding signal S3 is disable. Then the burst error form of entire data of a coded block is replaced with correction data and only code error data of a random error are replaced with the correction data. Error adaptive correction processing is applied to the data. Then a correction video decoding section 4 applies prescribed decoding processing to a corrected video coding signal S6 to decode image data S7.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to signal processing for decoding an image signal from a video coded signal whose information amount has been compressed by high efficiency coding, and more particularly to a code error which cannot be corrected in the video coded signal. The present invention relates to a method and apparatus for decoding an image signal, which is suitable for decoding an image with extremely little deterioration in image quality even when the above occurs.

[0002]

2. Description of the Related Art In recent years, with the progress of high-efficiency coding technology,
BACKGROUND ART Transmission and reception of an image by a video coded signal in which an information amount of an image is efficiently compressed to a fraction to a few tenths, recording to a storage medium, and the like are performed.

Generally, in a video coded signal in which the amount of information is compressed by high efficiency coding, a code error generated in a transmission line or a storage medium is widely propagated to its peripheral area in the decoding process and is noticeable on a decoded image. Degradation of image quality occurs. That is, there is a problem that the image quality of a high-efficiency coded video coded signal is significantly deteriorated due to a code error occurring in a transmission system or the like.

To avoid this, an error correction code for correcting a code error generated in a transmission system or a storage medium is usually added in the channel coding unit on the transmitting side, and the channel decoding unit on the receiving side. A process of correcting a code error by the error correction code is performed. As a result, in the video coded signal output from the channel decoding unit, most of the code errors generated in the transmission system and the storage medium system are corrected, but the correction capability of the error correction code is limited, so that the error of the correction cannot be corrected. Possible code errors may occur. Therefore, for a code error that cannot be corrected, error correction signal processing is performed in which the code error is replaced with a signal having a strong correlation.

Conventionally, this error correction processing replaces a signal in a region where a code error has propagated with a decoded image signal output from a video decoding unit with an image signal having strong correlation, or prior to video decoding, Two methods are known in which code error data of a video coded signal is replaced with coded data having a strong correlation.

However, in the former method, in a video-encoded signal having high compression efficiency, a region in which a code error propagates extends over a wide range, so that the deterioration of the image quality due to the correction process is conspicuous and becomes an annoying obstruction. On the other hand, in the latter method, replacement may be performed with coded data having no correlation, and an unpleasant disturbance due to this may occur in the decoded image.

[0007]

The correction of code errors according to the above-mentioned prior art is such that, in a high-efficiency coded video coded signal, the area to be replaced and corrected by the image signal becomes wide, and the coded data is replaced and corrected. There is a problem that the above deterioration propagates to a wide range, and there is a problem that an undesired deterioration of the image quality occurs due to the correction.

An object of the present invention is to provide a decoding method and apparatus for an image signal, which realizes correction of a high-efficiency coded video coded signal that cannot be error-corrected or code error is extremely small in image quality deterioration. To do.

[0009]

In order to achieve the above object, according to the present invention, coded data correction is performed by replacing code error data of a video coded signal in which error correction is impossible with correction data. A part is provided to classify random error and burst error in the form of code error data, and all coded blocks including code error data that cannot be corrected by random error and code error data that cannot be corrected by burst error. The first image decoding method that performs error-adaptive correction processing in which data is replaced with correction data having a strong correlation is adopted.

Further, the first correction section for replacing the code error data of the video coded signal, which cannot be corrected with the error, with the correction data, and the correction process, and the error propagation area of the decoded image signal with the correction image data. A second correction unit that replaces and corrects is provided to classify random errors and burst errors in the form of code error data. The first correction unit for burst errors and the second correction unit for random errors. The second image decoding method in which the correction processing is performed by the correction unit is adopted.

[0011]

First, an outline of image decoding according to the conventional technique will be described with reference to FIG. The signal VI from the transmission line or storage medium is
The channel decoding unit 1 performs a predetermined demodulation process and a code error correction process using an error correction code, and decodes the coded bit stream signal S1. This signal is input to the multiplexed video decoding unit 3 via the buffer 2, and the video coded signal S3 (composed of video coded data, coding parameter information, motion vector information, etc.) is decoded. Then, the video decoding unit 4 performs a predetermined decoding process to obtain the image data S4.
To decrypt. The video correction unit 5 performs a correction process of replacing a region of the image data S4 in which a code error propagates with correction data, for example, image data of the previous frame. The image post-processing unit 6 performs predetermined signal processing such as format conversion of image data and obtains a decoded image signal VO at its output. In this case, as described above, since the area in which the correction process is performed covers a wide range, an unpleasant deterioration in image quality occurs.

Next, the outline of the first image decoding method of the present invention will be described with reference to FIG. Different from the above-mentioned conventional technique, a coded data correction unit 7 is provided between the multiplexed video decoding unit 3 and the video decoding unit 4, and the coded error data of the video coded signal S3 in which the error correction is impossible is performed. Perform correction processing. It should be noted that the conventional technique for performing correction processing on a coded video signal has the same configuration as that shown in the figure, but the correction processing in the encoded data correction unit 7 is largely different from the conventional technology.

That is, in the prior art, regardless of the form of code error data in which error correction is impossible, only the code error data is replaced with the correction data, or all the data of the coding block including the code error data is corrected. Replace with data,
The correction process is performed. For this reason, in the former case, an unpleasant deterioration in image quality due to the correction data occurs due to burst errors in a form in which code error data that cannot be error-corrected concentrates. In the latter case, sporadic random errors cause deterioration of image quality such as annoying block distortion due to the correction data. Therefore, in any case, the correction process causes annoying image quality disturbance,
The quality of the decoded image deteriorates.

On the other hand, according to the present invention, by performing the error adaptive type correction process described below, it is possible to obtain a high quality decoded image with very little deterioration of the image quality occurring in the prior art. That is, according to the present invention, the form of the code error is classified into a burst error and a random error according to the occurrence status of the code error data in which the error correction of the video coded signal is impossible. Then, in the case of burst error, all the data of the encoded block is replaced with the correction data, and in the case of random error, only the code error data is replaced with the correction data, and an error adaptive correction process is performed. Table 1 shows an example of this error adaptive correction processing.

[0015]

[Table 1]

In any of the correction methods A, B, and C, when a burst error occurs, the data of the coded block at the same position in the previous frame is used as the correction data to perform the replacement correction.

On the other hand, in the case of random error, different correction processes are performed in each method. First, in the A method, the code error data of any of the video data part, motion vector information part, and coding parameter part of the video coded signal uses the data of the coded block at the same position in the previous frame as the correction data. Replace and modify. Further, in the B system, regarding the code error data of the video data part, the DC component of the transform coefficient is the average value of the adjacent coding blocks of the same frame, AC.
The component is substitution corrected using a zero signal for the correction data.
Further, in the C method, the DC of the conversion coefficient of the video data part
The component and the code error data of the motion vector information part are replaced and corrected by using the average value of adjacent coded blocks of the same frame as the correction data.

As described above, according to the present invention, for code error data for which error correction is impossible,
Error-adaptive correction processing is performed using correction data in which deterioration of image quality due to replacement correction is visually inconspicuous. For this reason,
It is possible to perform a correction process with extremely little deterioration in image quality as compared with the conventional technique. In the case of a random error, for example, it is possible to further finely control the correction method such as the A method for the still portion and the B and C methods for the moving image portion to further reduce the image quality deterioration.

In high-efficiency coding, orthogonal transform coding is widely used to efficiently remove redundancy in the spatial direction (horizontal, vertical) of an image. In this encoding, the data of the block (N pixels x N lines) is converted into the conversion matrix (N x
The transform coefficient generated by the matrix operation with N) is encoded and transmitted. On the receiving side, block data is decoded by matrix calculation with the inverse transformation matrix. Therefore, in the orthogonal transform coding, there is a correspondence between the code error of the transform coefficient and the error propagation area in the decoded data.

FIG. 3 is a diagram showing this correspondence by taking DCT (discrete cosine transform) coding as an example. FIG.
If the coefficient Xi, j of the DCT transform coefficient is the code error data, the shaded area in the decoded data of FIG. 9B becomes the propagation area of the code error in the IDCT conversion processing. Therefore, the correction processing for replacing the data in the shaded area with the data in the area adjacent to the error propagation area, for example, the pixel X of (K, J) is the pixel A of (K, J-1) and the pixel A of (K, J-1). J + 1) pixel B may be replaced with the average value.

The one that carries out the correction processing according to this idea is
This is a second image decoding method of the present invention, and its outline is shown in FIG. In this method, the first correction unit 8 and the second correction unit 9
And two types of modification processing units are provided. Then, the form thereof is classified into a burst error and a random error according to the generation status of code error data in which the error correction of the video coded signal S3 is impossible. Then, for a burst error, the first correction unit 8 performs a correction process. On the other hand, in the case of a random error, the code error of the video data section corresponding to the transform coefficient is corrected by the second correction section, and the code error of the motion vector information section and the coding parameter section is corrected by the first correction section. Table 2 shows an example of this correction processing.

[0022]

[Table 2]

In both of the correction methods 1 and 2, when the burst error occurs, the data of the coded block at the same position in the previous frame is used as the correction data to perform the replacement correction. In the case of random error, with respect to the video data part, the correction process of replacing the error propagation region of the coding block with the corrected image data generated based on the pixels adjacent to this region is performed by the second correction part in any method. To do. On the other hand, with respect to the motion vector information part, the first correction part performs a correction process of replacing with the data of the adjacent encoded block of the previous frame for the method 1 and the current frame for the method 2. In addition, with respect to the encoding parameter part, in either method, the first correction part performs a correction process of replacing with the data of the previous frame.

As described above, with respect to code error data that cannot be error-corrected, the correction data is selected by selecting the correction data in which the image quality deterioration due to the replacement correction is visually inconspicuous according to the occurrence situation. To do. Therefore, it is possible to realize the correction process with extremely little image quality deterioration as compared with the conventional technique.

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of the present invention will be described with reference to an overall block diagram shown in FIG. The present embodiment is suitable for performing decoding by the first image decoding method.

The channel decoding unit 1 performs a predetermined demodulation process on the transmission signal VI and a code error correction process using an error correction code. Then, it outputs the decoded coded bit stream signal S1 and the error flag signal E1 indicating the location of the code error that cannot be corrected. These signals are input to the buffer 2 at a constant bit rate.

The coded bit stream signal S2 and the error flag signal E2 read from the buffer 2 are separated by the multiplexed video decoding unit 3 into a video stream signal and an audio stream signal, video coded data of the video stream signal, and a code. The decoding parameter information and the motion vector information are decoded to generate the video coded signal S3 and the error data flag signal E3.

The coded data correction unit 7 detects the occurrence status of a code error that cannot be corrected by the error data flag signal E3, determines whether it is a burst error or a random error, and determines the type of each error. Error-adaptive correction processing is performed in which code error data or data of a coded block including code error data is replaced with correction data. Then, the modified video coded signal S6 is output.

The video decoding unit 4 performs a predetermined decoding process on the modified video coded signal S6 to obtain the image data S
Decrypt 7. Then, the image post-processing unit 6 performs predetermined signal processing such as format conversion of image data and outputs the decoded image signal VO.

Next, FIG. 6 shows an example of the configuration of the encoded data correction section 7 in this embodiment. Random error correction data generation unit 10, correction mode setting unit 11, and switch 1
2, a memory unit 13, and a memory control unit 14,
As shown in Table 1, error adaptive correction processing is performed.

The Ramdam error correction data generation unit 10 generates correction data from the data of adjacent coded blocks in the same frame. Further, the correction mode setting unit 11 sets, for each coding block, a video flag CED, a vector flag CEV, and a parameter flag CEP of the error data flag signal E3 (a signal of 1 for code error data and a signal of 0 for data without code error). The number of 1 is measured and the correction processing mode is set. For example, 0 is set to the uncorrected mode, 1 is set to the random error correction mode, and 2 or more is set to the burst error correction mode. Then, the control signal necessary for the switching operation of the switch 12 corresponding to these modes is generated.

The memory section 13 delays the input signal for one frame period of encoding and outputs the data of the encoded block at the same position one frame before as the correction data. The memory control unit 14 generates control signals necessary for the operation of the memory unit.

The switches 12-1 to 12-3 are connected to the terminal a in the uncorrected mode, and the video coded signal S3 (video coded data SVD, motion vector information SMV, coding parameter information SCP) is directly corrected video. The encoded signal S6 (video encoded data CSVD, motion vector information CSMV, encoded parameter information CSCP) is output. Further, in the burst error correction mode, the switches 12-1 to 12-1 are connected over the entire data period of the encoded block.
12-3 is connected to the terminal c and outputs the correction data from the memory unit 13 to realize the replacement correction by the data of the previous frame shown in Table 1.

On the other hand, in the random error correction mode, the switches 12-1 to 12-3 are connected to the terminal a during the period of data having no code error in the coding block, and the switch 12 in the A system is used during the period of code error data. -1 to 12-3 are terminals c, and in the B system, switch 12-1 is terminal b and switch 12
-2 and 12-3 are terminals c, and a switch 12-in the C method.
1, 12-2 are connected to the terminal b, and the switch 12-3 is connected to the terminal c to realize the replacement correction shown in Table 1.

Since the other block units in this embodiment can be configured in the same manner as in the conventional decoding device, the description thereof will be omitted.

As described above, according to this embodiment, for coded data that cannot be error-corrected by the video coded signal,
An image signal decoding device that performs error correction processing with extremely little deterioration in image quality can be realized. Therefore, a remarkable effect can be obtained in improving the image quality of the video-encoded image.

Next, regarding the second embodiment of the present invention,
This will be described with reference to the overall block diagram shown in FIG. The present embodiment is suitable for performing decoding by the second image decoding method.

The transmission signal VI is subjected to a predetermined demodulation process and a code error correction process by an error correction code in the channel decoding unit 1, and the decoded coded bit stream signal S1 is obtained.
And an error flag signal E1 indicating the location of a code error that cannot be corrected. These signals are input to the buffer 2 at a constant bit rate.

The coded bit stream signal S2 and the error flag signal E2 read out from the buffer 2 are input to the multiplexed video decoding unit 3, the video stream signal and the audio stream signal are separated, and the video coded data of the video stream signal is input. , The encoding parameter information and the motion vector information are decoded to generate a video encoded signal S3 and an error data flag signal E3.

The first correction section 8 receives the error data flag signal E
From 3, the occurrence status of a code error that cannot be corrected is detected to determine whether it is a burst error or a random error. Then, for the burst error, as shown in Table 2, a correction process of replacing the data of the coding block including the code error data with the correction data is performed, and the corrected video coded signal S8 is output. For random errors, it outputs a signal ECA indicating the location of code error data.

The video decoding unit 4 performs a predetermined decoding process on the modified video coded signal S8 to generate image signal data S9. Then, the second correction unit 9 performs a correction process on the decoded image data S9 by detecting an error-propagating region from the signal ECA and replacing the image data in this region with the corrected image data. Using the corrected image decoded data, the video decoding unit 4 continues the decoding process to decode the image data S7.

The image post-processing unit 6 performs predetermined signal processing such as format conversion of image data and outputs the decoded image signal VO.

Next, FIG. 8 shows an example of the configuration of the first correction section 8 in this embodiment. The random error correction data generation unit 10, the correction mode setting unit 11, the switch 12, the memory unit 13, the memory control unit 14, and the video data correction mode setting unit 21 are included.

The random error correction data generation section 10 generates correction data based on the data of adjacent coded blocks of the same frame. In addition, the correction mode setting unit 11
Is 1 of the vector flag CEV and the parameter flag CEP of the error data flag signal E3 (1 for code error data, 0 for data without code error) for each coding block.
The number of is measured, and the correction processing mode is set. For example, the number 0 is set to the uncorrected mode, the number 1 is set to the random error correction mode, and the number 2 or more is set to the burst error correction mode. And the switch 12 corresponding to these modes
The control signal necessary for the switching operation of is generated.

The memory unit 13 delays the input signal for one frame period of encoding and outputs the data of the encoded block at the same position one frame before as the correction data. Also,
The memory control unit 14 generates control signals necessary for the operation of these memory units.

The video data correction mode setting unit 21 sets 1 of the video flag CED of the error data flag signal E3 (1 for code error data and 0 for data without code error).
The number of is measured, and the correction processing mode is set. For example, the number 0 is set to the uncorrected mode, the number 1 is set to the random error correction mode, and the number 2 or more is set to the burst error correction mode. And the switch 12 corresponding to these modes
The control signal necessary for the switching operation of -4 is generated. Further, in the random error correction mode, the signal ECA of which the location of the code error data is 1, for example, and 0 otherwise is output.

In the uncorrected mode, the switches 12-2 to 1
2-4 are both connected to the terminal a, and the video coded signal S
3 (video coded data SVD, motion vector information SM
V, the coding parameter information SCP) is output as it is as the modified video coded signal S8 (video coded data CSVD, motion vector information CSMV, coding parameter information CSCP).

Further, in the burst error correction mode, all the switches 12-2 to 12-4 are connected to the terminal c and output the correction data from the memory section 13 over the entire data period of the encoded block. The replacement correction by the data of the previous frame shown in Table 2 is realized.

On the other hand, in the random error correction mode, the switch 12-4 is connected to the terminal a and outputs the video encoded data SVD including the code error data as it is to the signal CSVD. The correction of the code error data is executed by the second correction unit 9 described later. Also, the switches 12-2, 1
2-3 is connected to the terminal a during the data period having no code error. Then, in the period of data of code error, in the method 1, all are connected to the terminal c, in the method 2, the switch 12-2 is connected to the terminal b and the terminal 12-3 is connected to the terminal c, and the replacement correction shown in Table 2 is realized.

Next, an example of the configuration of the second correction section 9 is shown in FIG. Note that this configuration is for the case where the video coded signal is coded by combining DCT coding and motion compensation interframe predictive coding.

First, the video decoding unit 4 will be described. The encoded video data CSVD is input to the IVLC unit 15 and is subjected to a process of decoding the variable length code into the original fixed length code. The IQ unit 16 performs inverse quantization processing with the quantization characteristic specified by the encoding parameter information CSCP, and decodes the DCT transform coefficient. The IDCT unit 17 performs matrix operation on the DCT transform coefficient and the inverse DCT matrix. Then, the image data is decoded in the intra-frame coding mode, and the prediction error data signal PE is decoded in the inter-frame coding mode. Adder 18
Adds the signal PE and the prediction signal PS to generate image data in the interframe coding mode. Switch 19
In accordance with the coding mode designated by the coding parameter information CSCP, is connected to the terminal a for intra-frame coding and the terminal b for inter-frame coding, and obtains the decoded image data S9 at its output. This signal is subjected to a correction process in which a region in which a code error propagates is replaced with corrected image data in a second correcting section 9 described later, and the image data S7 is decoded. The MC prediction signal generation unit 20 performs motion compensation signal processing on the image data S7 based on the motion vector of the motion vector information CSVD to generate a prediction signal PS.

The second correction section 9 comprises a memory section 22, a correction data generation section 23, an error propagation area detection section 24, and a switch 33. Then, the error propagation area detection unit 2
4, the area where the code error data propagates as shown in FIG. 3B is calculated from the information on the location of the code error data obtained from the signal ECA (hatched portion in the figure), and in this area, The switch 33 generates a control signal for performing the operation of selecting the corrected image data of the terminal b. Image decoded data S9
Is input to the memory unit 22, and the image decoded data delayed by a predetermined time from one of the outputs is supplied to the terminal a. The other output is input to the correction data generation unit 23, and
As shown in (b), the corrected image data is generated based on the image data adjacent to the error propagation area and is supplied to the terminal b. Then, at the output of the switch 33, the image data S7 in which the image data in the error propagation area is replaced with the corrected image data having a very strong correlation is obtained.

Since the other block units in this embodiment can be constructed in the same manner as in the conventional decoding device, the description thereof will be omitted.

As described above, according to this embodiment, for coded data that cannot be error-corrected by the video coded signal,
It is possible to realize an image signal decoding device that performs error correction processing with extremely little deterioration in image quality. Then, a remarkable effect can be obtained in improving the image quality of the video coded image.

Next, the third embodiment of the present invention will be described.
This will be described with reference to the block diagram shown in FIG. The present embodiment is suitable for decoding the coded video signal recorded in the storage medium by the first image decoding method.

The reproduced signal VP from the storage medium is subjected to predetermined demodulation processing such as waveform equalization and maximum likelihood decoding in the channel decoding section 25 to decode the recording data stream signal S11. The deinterleave processing unit 26 performs a predetermined time-series rearrangement process and decodes the encoded bitstream signal S12. The code error correction unit 27 performs a code error correction process using an error correction code and decodes the video coded signal S13. Further, it outputs an error flag signal E4 consisting of a signal of 1 for code error data that cannot be error-corrected and a signal of 0 for data without code error.

The coded data correction unit 28 detects the occurrence status of a code error that cannot be error-corrected by the error flag signal E4, and classifies it into a burst error correction mode and a random error correction mode. Then, for the video coded signal S13, an error-adaptive correction process of replacing the coded block by the correction data in the burst error correction mode and replacing the code error data by the correction data in the random error correction mode is performed. Then, the modified video coded signal S14 is generated.

The video decoding unit 29 performs a predetermined decoding process on the modified video coded signal S14 to decode the image data S15. Then, the image post-processing unit 30 performs predetermined signal processing such as format conversion of image data and obtains the decoded image signal VO at its output.

The coded data correction unit 28 in this embodiment
Can be realized by the configuration shown in FIG. 6 and the signal processing shown in Table 1. Further, the other blocks can be easily realized in the same manner as the conventional technique. Therefore, the description regarding these is omitted.

As described above, according to the present embodiment, an image signal for which an error correction process with extremely little deterioration in image quality is performed with respect to a code error in which the error correction of the video coded signal recorded in the storage medium is impossible. A recording device can be realized. Then, it is possible to obtain a remarkable effect in improving the image quality of images such as package media.

Next, regarding the fourth embodiment of the present invention,
This will be described with reference to the block diagram shown in FIG. The present embodiment is suitable for decoding the coded video signal recorded in the storage medium by the second image decoding method.

The reproduced signal VP from the storage medium is subjected to predetermined demodulation processing such as waveform equalization and maximum likelihood decoding in the channel decoding unit 25 to decode the recording data stream signal S11. The deinterleave processing unit 26 performs a predetermined time-series rearrangement process and decodes the coded bit stream signal S12. The code error correction unit 27 performs a code error correction process using an error correction code and decodes the video coded signal S13. Further, it outputs an error flag signal E4 consisting of a signal of 1 for code error data that cannot be error-corrected and a signal of 0 for data without code error.

The first correction section 31 detects the occurrence of a code error that cannot be corrected by the error flag signal E4, and classifies it into a burst error correction mode and a random error correction mode. Then, in the burst error correction mode, as shown in Table 2, a correction process of replacing the coding block with the correction data is performed, and the corrected video coded signal S16 is output. In the random error correction mode, the signal ECA indicating the location of the code error data is output.

The video decoding unit 29 performs a predetermined demodulation process on the modified video coded signal S16 to generate image decoded data S18. This signal is sent to the second correction unit 32.
Then, a correction process is performed to replace the error propagating area detected from the signal ECA with the correction image data. Then, by using this corrected image decoded data, the video decoding unit 29
Continues the decoding process to decode the image data S17, and the image post-processing unit 30 performs predetermined signal processing such as format conversion of the image data and obtains the decoded image signal VO at its output.

Next, the second correction section 32 in this embodiment.
FIG. 12 shows an example of the configuration. Note that this configuration is for the case where the video coded signal is DCT coded within a frame.

First, the video decoding unit 29 will be described. The video coded data CSVD of the modified video coded signal S16 is input to the IVLC unit 15 and is subjected to processing of decoding the variable length code into the original fixed length code. The IQ unit 16 is
The inverse quantization process is performed with the quantizing property specified by the coding parameter information CSCP, and the DCT transform coefficient is decoded. IDCT part 1
Reference numeral 7 performs a matrix operation of the DCT transform coefficient and the inverse DCT matrix to decode the image decoded data S18.

On the other hand, the second correction unit 32 includes the memory unit 22.
A correction data generation unit 23 and an error propagation area detection unit 24
And a switch 33. Then, the error propagation area detection unit 24, based on the information on the location of the code error data of the signal ECA, as shown in FIG. ) Is calculated, and the control signal is generated so that the switch 33 selects the corrected image data of the terminal b in this area.

The image decoded data S18 is input to the memory unit 22, and the image decoded data delayed by a predetermined time from one of the outputs is supplied to the terminal a. The other output is input to the correction data generation unit 23, and as shown in FIG. 3B, correction image data is generated based on the image data adjacent to the error propagation area and is supplied to the terminal b. . And the switch 33
In the output of, the image data S17 is obtained by replacing the image data in the error propagation area with the corrected image data having a very strong correlation.

It should be noted that the first correction section 31 in the present embodiment.
Can be realized by the configuration shown in FIG. 8 and the signal processing shown in Table 2. Further, the other blocks can be easily realized in the same manner as the conventional technique. Therefore, the description regarding these is omitted.

As described above, according to the present embodiment, an image signal for which error correction processing with extremely little deterioration in image quality is performed for a code error that cannot be error-corrected in a video coded signal recorded in a storage medium. A recording device can be realized. Then, it is possible to obtain a remarkable effect in improving the image quality of images such as package media.

In the description of the embodiment, the DCT coding is taken as an example of the orthogonal transform coding, but it is obvious that the present invention can be applied to, for example, Hadamard transform coding.

Video coded in various forms such as intra-frame orthogonal transform coding, or combination of intra-frame orthogonal transform coding and inter-frame predictive coding, or motion-compensated inter-frame predictive coding. It is obvious that the decoding method according to the present invention can be applied to a coded signal.

[0073]

According to the present invention, even if an uncorrectable coding error occurs in a video coded signal, it is possible to decode an image with extremely little deterioration in image quality. Therefore, it is possible to obtain a remarkable effect in improving the image quality of images in the fields of moving image storage media, digital broadcasting, digital CATV, TV phones, TV conferences, etc., in which information is compressed and transmitted and received.

[Brief description of drawings]

FIG. 1 is a block diagram of a first image decoding according to the present invention.

FIG. 2 is a block diagram of image decoding according to the related art.

FIG. 3 is an explanatory diagram of code errors of DCT coefficients and decoded data.

FIG. 4 is an explanatory diagram of second image decoding according to the present invention.

FIG. 5 is an overall block diagram of a first embodiment of the present invention.

FIG. 6 is a block diagram of an encoded data correction unit according to the first embodiment.

FIG. 7 is a block diagram of a second embodiment of the present invention.

FIG. 8 is a block diagram of a first correction unit according to the second embodiment.

FIG. 9 is a block diagram of a second correction unit according to the second embodiment.

FIG. 10 is a block diagram of a third embodiment of the present invention.

FIG. 11 is a block diagram of a fourth embodiment of the present invention.

FIG. 12 is a block diagram of a second correction unit of the fourth embodiment.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Channel decoding part, 2 ... Buffer, 3 ... Multiplexed video decoding part, 4 ... Video decoding part, 5 ... Video correction part, 6
... Image post-processing section.

Claims (12)

[Claims] 1. A method of decoding an image signal for decoding a video coded signal, the information quantity of which has been compressed by a high efficiency coding technique, wherein a code error in which the video coded signal cannot be corrected prior to video decoding. A coded data correction unit that replaces data with correction data to perform correction processing is provided, and the coded data correction unit classifies random errors and burst errors in the form of code error data that cannot be corrected, and The code error data that cannot be corrected by an error is replaced with the correction data, and all the data of the coding block (N × N, N is an integer) having the code error data that cannot be corrected by the burst error are corrected data. An image signal decoding method characterized by performing error-adaptive correction processing for replacement. 2. The correction data used in the burst error correction process according to claim 1, wherein an image generated by the encoded data of the encoded block at the same position at least one frame before the encoded block of the current frame. Signal decoding method. 3. The image data generated according to claim 1 or 2, wherein the correction data used for the random error correction processing is the coded data of the adjacent coded block of the same frame with respect to the coded block of the current frame. Decryption method. 4. The correction coefficient data correction process according to claim 1 or 2, wherein the DC component is generated by DC component data of adjacent coding blocks of the same frame, and the AC component is generated by zero data. A method for decoding an image signal to be replaced with data. 5. A method of decoding an image signal for decoding a video coded signal, the information quantity of which is compressed by a high efficiency coding technique, wherein a code error in which the video coded signal cannot be corrected prior to the video decoding. A first correction unit that replaces data with correction data to perform correction processing, and a corrected image of an error propagation area of code error data in which the video coded signal cannot be corrected in the decoded image signal of video decoding. A second correction unit that replaces with data to perform correction processing is provided, and the form of code error data in which the video coded signal cannot be corrected is classified into a random error and a burst error. A method of decoding an image signal, wherein correction processing is performed by a correction section, and random errors are corrected by the second correction section. 6. The correction process in the first correction section according to claim 5, wherein all the data of the coded blocks having code error data that cannot be corrected are coded blocks at the same position at least one frame before. A decoding method of an image signal for performing a correction process of replacing with correction data generated by the encoded data of. 7. A decoding method of an image signal according to claim 5, wherein the corrected image data used for the correction processing in the second correction section is generated based on an image signal in an area adjacent to an error-propagating area. . 8. The coded video signal according to claim 1, 2, 3, 4, 5, 6 or 7, wherein the information amount is compressed by a high efficiency coding technique by orthogonal transform coding within a frame. Image signal decoding method. 9. The video coded signal according to claim 1, 2, 3, 4, 5, 6 or 7, wherein the video coded signal is highly efficient in DCT (discrete cosine transform) coding and motion compensation interframe predictive coding. A decoding method of an image signal which is a signal in which the amount of information is compressed by an encoding technique. 10. Claims 1, 2, 3, 4, 5, 6 or 7
In H.2, the video coded signal is an international standard.
61, a decoding method of an image signal which is a signal based on MPEG-1 and MPEG-2 video coding. 11. A method according to claim 1, 2, 3, 4, 5, 6 or 7.
In which a video signal is recorded in a storage medium or reproduced from the storage medium, and an image signal for decoding the reproduced video coded signal by the image signal decoding method is recorded. apparatus. 12. A video signal recording apparatus according to claim 11, wherein the video coded signal is the signal according to claim 8, 9, or 10.