JPH10136369A - Decoder for digital moving image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a decoder for digital moving image in which the occurrence of disturbance of an image due to decoding error is prevented by setting a concealment start position to an image block before an image block from which an error is detected. SOLUTION: A decoding means 120 decodes a code stream of digital moving image to obtain image information, decode position information, a motion vector and error information and motion compensation processing is applied to the image information at a position designated by the decode position information by using the motion vector and error concealment processing is applied to the image information at a position designated by the error information. The decoder for digital moving image as above is provided with an error concealment position decision means 145 that provides an output of error concealment position information denoting a start position of the error concealment processing based on the error information and the decode position information and with a decode image position decision means 140 that decides a decode position based on the error concealment position information and the decode position information. Thus, error concealment is attained for a position at which a true error is produced.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital moving picture decoding apparatus used for transmitting a moving picture of a digital signal via a transmission line or storing the moving picture in a storage device.

[0002]

2. Description of the Related Art In recent years, with the progress of the information-oriented society, there has been an increasing demand for transmitting moving images to other people over time and distance barriers. At the same time, it has become possible to record and reproduce moving images with a recording device and to transmit data over long distances using a communication network, and a situation is being formed in which the above-mentioned demand can be realized.

[0003] Further, in the field of broadcasting as well as in the field of communication, transmission and coding methods using digital technology are being adopted, and the coding and decoding of moving images is more efficient and higher quality. It is hoped that this can be realized.

Since a digital signal of a moving image has a much larger amount of information than an audio signal, it is necessary to use a high-efficiency coding technique to efficiently record and transmit the digital signal. Encoding and decoding devices capable of withstanding various requirements have already been prototyped.

[0005] However, the digital video which has been highly coded has a feature that, since a large amount of information is compressed per unit capacity, if noise is mixed in the transmission path, the video after decoding is significantly deteriorated. .

For this reason, an error correction code may be added to a code string after high-efficiency coding to improve error resilience. However, even in such a case, if there are many errors in the transmission path, it is completely lost. It becomes difficult to decode a moving image having no error and no deterioration.

In order to visually alleviate the deterioration of the decoded moving image, there is a method of concealing an image area of an error occurrence portion of a highly efficient coded digital moving image by using an already decoded image. . As a digital video decoding apparatus using such a method, an international standard (ISO /
IEC13818-2) "Information Technology Generic Coding of Moving Pictures and Associated Audio for Digital Information"
echnology-Generic coding of moving pictures and
associated audio for digital information ".

Hereinafter, an example of a conventional moving picture decoding apparatus using the above-described method will be described with reference to FIG. The code string of the digital moving image input to the input terminal 100 is stored in the buffer 11.
It is buffered once at 0 and input to the decoding means 120.

The decoding means 120 includes decoding position information indicating a screen position of an image block constituting a decoded image, error information indicating whether an error has been detected during decoding of an input code string, motion vectors, DCT coefficients, and the like. The image information is decoded and output.

[0010] The decoding screen position determining means 140 determines the screen position of the image block to be decoded based on the decoding position information output from the decoding means 120, and notifies the motion compensating means 200. Further, the decoded motion vector is temporarily stored in the motion vector memory 190 and then supplied to the motion compensating means 200.

If no error is detected by the decoding means 120, the image information is inversely quantized by the inverse quantization means 150 and the inverse DCT (Discrete
Orthogonal transform decoding such as Co-sin Transform) is performed.
The result is added to the image data of the image memory 180 for storing the decoded image data by the adder 170 and stored in the image memory 180.

The decoded image data stored in the image memory 180 is subjected to motion compensation processing by the motion compensating means 200 using the motion vector. The decoded image data is stored in the image memory 1 in the order of output.
80, and output from the output terminal 300.

If an error is detected by the decoding means 120, the decoding means 20 discards the code until the position immediately before the next error-free code sequence position in the coded sequence of the digital moving image. Further, the decoding screen position determining means 140 performs a motion so as to extract the motion vector held by the image block one immediately above the screen position of the image block in which the error has occurred in the image currently being decoded from the motion vector memory 190. An instruction is given to the vector memory 190, and the motion vector is given to the motion compensating means 200.

Using this motion vector, the motion compensating means 200 performs motion compensation for the decoded image data, for example, the image data at the same position on the previous screen stored in the image memory, and The value of the block is stored in the image memory 180, and error concealment is performed on the image block in which an error is detected.

This concealment processing is performed up to immediately before the image block position indicated by the next error-free code string position. The error concealment processing when the above-described conventional digital video decoding apparatus is used will be described with reference to FIG. That is, it is assumed that the currently decoded image 50 includes the image block 21 in which the error is detected, and the image blocks 10, 11, 12, 2,
It is assumed that 0, 22, 30, 31, and 32 are adjacent to the image block 21 in which the error has been detected. Further, the arrows in the figure indicate the above image blocks 10, 11, 12, 20, 2,
The motion vectors held by 2, 30, 31, and 32 are schematically shown.

In the above-described processing by the conventional digital video decoding apparatus, the image block 21 in which an error is detected is used.
The motion compensation processing is performed using the motion vector held by the image block 11 on the upper side of, and the error concealment processing is performed.

[0017]

In the above error detection by the decoding means 120, there is a case where a true error is present in the image block 21 in which the error is detected. In some cases, an error is detected in the image block 21 subsequent to the data included in the image block 21.

In this case, it is not sufficient to pay attention only to the error of the image block 21, and it is necessary to perform an error concealment process on an image block containing a true error before that. However, in the above-described conventional configuration, only the image block 21 in which an error has been detected is subjected to error concealment processing, motion compensation is performed on the decoded image data, and the result is used as the value of the decoded image block. If the error occurrence position is before the image block position 21 where the error is actually detected by the decoding device, the processing for the error is performed on the image blocks from the screen position to the image block 21. Is not performed, and there is a defect that significant image disturbance occurs even though the error concealment processing is performed.

The present invention has been proposed in view of the above-mentioned conventional circumstances, and an image block prior to an image block in which an error has been detected is set as a concealment start position, thereby generating image disorder due to a decoding error. It is an object of the present invention to provide a digital video decoding device capable of preventing the problem.

[0020]

In order to solve the above-mentioned problems, the present invention employs the following means. That is, a code string of a digital moving image is decoded by a decoding means 120 and the image information is decoded. The position information, the motion vector, and further error information are obtained, and a motion compensation process is performed on the image information at the position specified by the decoded position information using the motion vector, and the image information at the position indicated by the error information is obtained. An error concealment position determining unit 145 that outputs error concealment position information indicating a start position of the error concealment process based on the decode position information and the error information. Decoding position determining means 140 for determining a decoding position based on the error concealment position information and the decoding position information. It is.

Thus, the error concealment start position can be traced back to the position of the image block where the error is detected, and the error concealment including the true error position can be performed.

As the error concealment start position, the immediately preceding position in the decoding order of the image block position in which the error is detected, the first image block position of a specific image block group to which the image block in which the error is detected belongs, or the error position. The image block position at the left start end of the column to which the detected image block belongs can be set.

[0023]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The specific structure of the present invention will be described below with reference to FIGS. (Embodiment 1) FIG. 1 is a block diagram showing an embodiment of a digital video decoding apparatus according to the present invention. In FIG. 1, a code string of a digital moving image input to an input terminal 100 is temporarily buffered in a buffer 110,
Input to 0. In the decoding means 120, the decoding position information, error information, motion vector and DCT
The point that image information such as coefficients is decoded and output is the same as the above-described conventional configuration and operation.

The error information indicating the error position output from the decoding means 120 is input to the error concealment position determination means 145, and the error concealment position determination means 145 outputs the error information and the output from the decoding means 120. The error concealment start position is determined based on the decoded position information.

The error concealment position information obtained by the error concealment position determination means 145 in this way is
0, the decoding screen position determining means 140 determines the position of the image block to be decoded based on the decoding position information output from the decoding means 120 and the error concealment position information, At the same time, the motion vector memory 190 is also notified.

The motion vector decoded by the decoding means 120 is stored in the motion vector memory 1 as in the prior art.
90 and is supplied to the motion compensation means 200.
If no error is detected by the decoding means 120,
As in the conventional case, the image information is inversely quantized by the inverse quantization means 150, and orthogonal transform decoding such as inverse DCT is performed by the orthogonal transform decoding means 160. The result is added to the image data of the image memory 180 for storing the decoded image data by the adder 170 and stored in the image memory 180.
The image memory 180 stores the decoded image data of the screen.

The decoded image data stored in the image memory 180 is read out from the image memory 180 in the order in which the decoded image is to be output, and the output terminal 300
Output from

If an error is detected by the decoding means 120, the decoding means 120 discards the code until the next error-free code sequence position in the coded sequence of the digital moving image.

The error concealment position determining means 145 receives the error information indicating the error position detected by the decoding means 120 and the decoding position information at the same time as the above error detection. Based on these two pieces of information, error concealment position information indicating the start position of error concealment is output so that decoding is performed again from the immediately left image block position immediately before the image block currently being decoded in the decoding order.

The error concealment position information thus obtained is input to the decoding screen position determining means 140, and the decoding position information is further input to the decoding screen position determining means 140. Based on these two pieces of information, the decode screen position determining means 140 determines the next image block position corresponding to the designated image block position, that is, the image block to the left of the screen position of the erroneous image block. Instructs the motion vector memory 190 to fetch the motion vector held by. As a result, the motion vector memory 190 stores the motion compensating means 20
Give the motion vector to 0.

In this manner, the concealment process is similarly continued until immediately before the image block position indicated by the next error-free code string position. The motion compensating means 200 motion-compensates the decoded image data (for example, the image data of the image block at the corresponding position in the previous screen), stores it in the image memory 180 as the value of the decoded image block, and Perform concealment.

FIG. 2 illustrates the above-described error concealment process, in which the motion vectors are indicated by arrows. If the currently decoded image 50 includes an image block 21 in which an error is detected, the left image block 20
Is the error concealment start position, the motion is compensated using the motion vector 10 of the image block above the image block 20, the error concealment process of the image block 20 is performed, and the motion vector 1
1 to perform motion concealment processing on the image block 21 by performing motion compensation.

As described above, according to the digital video decoding apparatus of the present invention, since the error concealment process is performed from the image block position immediately before the image block position 21 where the error is detected, a true error is detected immediately before. In the case of an image block, it is possible to prevent significant image disturbance that may occur when decoding the image block, and to quickly complete the error concealment process.

(Embodiment 2) In the first embodiment, the error concealment position at which the error concealment process is started, which is determined by the error concealment position determination means 145, is the image block to the left of the image block position where the error is detected. However, the error concealment process can be started from the first image block position of the image block group to which the image block in which the error is detected belongs.

That is, in order to increase the compression density of image information, one or more image block groups continuous on the screen are
Often encoded together, slice or GOB
(Group of blocks). In this method, each block group is divided by using a unique start code indicating the head of the image block group to be collectively encoded.

Using this method, the start position of the error concealment processing can be set to the first image block of the group of image blocks to be encoded. In FIG. 1, the operation other than the error concealment position determining means 145 is the same as that of the first embodiment, and therefore the description is omitted.

The error concealment position determining means 145 first detects the first image block 40 of the image block group 60 to which the image block 21 in which the error is detected belongs as shown in FIG. Error concealment position information to be output.

Normally, the start position of the image block group 60 in the coded sequence is a minimum unit that can be easily specified because it is a unique start code. Therefore, even in the case of a digital video decoding apparatus employing the variable-length code decoding method, if an image block including the above-mentioned unique start code is present in the variable-length code, the image block becomes an image block. Group 6
This is the first image block 40 of 0.

When an error is detected by the decoding means 120, the decoding process has already reached the screen position of the image block 21, but the error concealment process is performed by going back to the image block 40. That is, the error concealment processing of the image block 40 is motion-compensated using the motion vector held by the image block 1, and the same applies to the image block 22 immediately before the image block position indicated by the error-free code string position. The concealment process is continued.

As described above, according to the digital video decoding apparatus of the present invention, the image block 2 in which an error is detected
In order to perform the error concealment process from the first image block position of the image block group to which the image block group 1 belongs, if a true error exists in the image block group 60, the image having no error located before the image block group 60 No extra error concealment processing is performed on the blocks, and significant image disturbance that may occur when decoding an erroneous image block can be prevented, so that effective error concealment processing can be performed.

(Embodiment 3) The error concealment processing start position may be further set to the left start end of the image block sequence to which the image block 21 in which the error has been detected belongs as follows.

In FIG. 1, error concealment position determining means 14
Operations other than 5 are the same as those in the first embodiment, and a description thereof will not be repeated. In FIG. 4, when an error is detected by the decoding unit 120, the error concealment position determination unit 145 outputs error concealment position information indicating the position of the image block at the left end of each column as the error concealment processing start position. .

At this time, the decoding process has already reached the screen position of the image block 21. However, the decoding position determining means 140 obtains the error concealment position information as described above, thereby changing the decoding position to the image in which the error is detected. The error concealment process is performed by going back to the position of the leftmost image block of the image block sequence to which the block 21 belongs. That is, in the error concealment processing of the image block 45, motion compensation is performed using the motion vector held by the image block 2 on the image block 45, and similarly, until immediately before the image block position indicated by the code string position having no error. Concealment processing is continued.

As described above, according to the digital video decoding apparatus of the present invention, the image block 2 in which an error is detected
Since the error concealment processing is performed from the leftmost image block of the image block sequence to which 1 belongs, if a true error is between the image blocks 21 where the error is detected from the left end of the screen, the start position of the image block group It is not necessary to hold information on the image block, and it is possible to prevent significant image distortion that may occur when decoding an erroneous image block.

(Others) In all of the embodiments described above, examples have been described in which DCT is used as orthogonal transform decoding. However, the present invention is not limited to this. Discrete wavelet transform, discrete sine transform Is also applicable.

Further, in the above-described embodiment, a case has been described in which the motion vector previously stored in the motion vector memory 190 is used as the error concealment process, and the decoded image is motion-compensated for concealment. However, the present invention is also applicable to a case where pixel information of a previous screen at the same screen position as an image block to be hidden in a decoded image is simply copied without using a motion vector.

That is, since the image information of the previous screen or the screen before the previous screen (or the rear screen) is normally stored in the image memory 180, the image position information output from the decode screen position determining means 140 in FIG. Based on this, the image block of the previous screen corresponding to the image block is read out from the image memory 120 to the image correcting means (corresponding to the motion compensating means 200 in FIG. 1), and the image is corrected. At this time, no motion vector is used.

Similarly, without using a motion vector, image information interpolated from pixel values of an image block adjacent to an image block to be concealed can be used. In this case, as shown in FIG. 5, the interpolation unit 210 stores the image information around the image block to be subjected to the concealment processing in the image memory 120.
And performs interpolation processing. Also at this time, the motion compensation means 200 does not substantially operate.

When using the image information that is currently being decoded, the adjacent, but upper, upper left,
Each image information on the left side is used, and when using the image data of the previous screen, for example, upper, lower, left and right image information adjacent to each other can be used as image blocks. As the interpolation method, the average value of the image information of each adjacent image block can be used.

[0050]

As described above, according to the present invention, by performing the error concealment process from the position before the image block position where the error is detected, the position of the image block where the true error is detected can be improved. In this case, it is possible to prevent a significant image disorder that may occur when decoding the image block, and to quickly complete the error concealment process.

The position before the image block position where the error was detected includes the immediately preceding position image block, the image block at the head position of the image block group to which the image block corresponding to the position where the error was detected, and the error position. The image block at the position of the left start end of the image block sequence to which the detected image block belongs can be used.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a digital video decoding apparatus according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram of a motion vector used for error concealment processing according to the first embodiment of the present invention.

FIG. 3 is an explanatory diagram of a motion vector used for error concealment processing in Embodiment 2 of the present invention.

FIG. 4 is an explanatory diagram of a motion vector used for error concealment processing in Embodiment 3 of the present invention.

FIG. 5 is a block diagram showing a configuration of a digital video decoding apparatus according to another embodiment of the present invention.

FIG. 6 is a block diagram showing a configuration of a conventional digital video decoding apparatus.

[Explanation of symbols]

 Reference Signs List 120 decoding means 140 decoding screen position determination means 145 error concealment position determination means

Claims (4)

[Claims] An image information, a decoded position information, a motion vector and further error information are obtained by decoding a code sequence of a digital moving image by a decoding means, and the image information at a position designated by the decoded position information is obtained. A digital video decoding device that performs a motion compensation process using a motion vector and performs an error concealment process on the image information at the position indicated by the error information, based on the decoded position information and the error information, Error concealment position determining means for outputting error concealment position information indicating a start position of error concealment processing; and decode screen position determination means for determining a decoding position based on the error concealment position information and the decode position information. Characteristic digital video decoding device. 2. The error concealment position determining means outputs error concealment position information indicating, as an error concealment start position, an immediately preceding position in the decoding order of an image block position in which an error has been detected when error information is in an error detection state. 2. The digital moving picture decoding apparatus according to claim 1, wherein: 3. An error concealment position determining means, when error information is in an error detection state, an error indicating a first image block position of a specific image block group to which an image block in which an error has been detected belongs as an error concealment start position. 2. The digital moving picture decoding apparatus according to claim 1, wherein the decoding apparatus outputs concealment position information. 4. The error concealment position determining means, wherein when the error information is in an error detection state, the error concealment position information indicating a left starting end image block position of a column to which the image block in which the error is detected belongs as an error concealment start position. 2. The digital moving picture decoding apparatus according to claim 1, wherein