JPH06276509A - Error correction method in high efficiency coding transmission and its transmitter-receiver - Google Patents

Abstract

PURPOSE:To obtain a reproduced picture with excellent quality even when a transmission error takes place and a normal picture signal is lost in video digital transmission. CONSTITUTION:A sender side sends an important component of a video signal as additional information before coding coefficient information. When an error check section 102 of a receiver side checks an error in coded coefficient information, a motion compensation section 112 compensates motion of an in-frame DC component sent already and extracts the result via a switch 108 as a decoded signal of a picture element block.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an error correction method in high-efficiency coded transmission in which a digital video signal is encoded and transmitted for each pixel block, and the encoded signal is received and decoded, and a transmitter / receiver thereof. .

[0002]

2. Description of the Related Art Conventionally, in a system in which a digital video signal is encoded and transmitted for each pixel block, a pixel block in an image frame in which a normal signal is lost due to a transmission error on the receiving side has the same pixel as that block. An error correction method is known in which data is replaced with data of a block that is one frame before or one frame later, or several frames before or several frames later, corresponding to a position.

FIG. 5 is a block diagram of a moving picture decoding apparatus in which, when a transmission error occurs in a pixel block, it is replaced with data of a block corresponding to the same pixel position one frame before.

In the figure, the coded data introduced to the terminal 100 is temporarily stored in the buffer memory unit 601.
It is input to the error detection unit 602. The encoded data includes a luminance (Y) signal and a C (color) signal as encoding coefficients, and additional information such as block synchronization / address information, block mode information (intra-frame / inter-frame processing mode signal), and motion. Vector information is multiplexed. The signal passed through the error detector 602 is input to the data separator 603.

In the data separation unit 603, the coded data is separated into sync / address information, mode information, motion vector and coding coefficient. The coding coefficient is supplied to the adder 607 via the variable length decoding unit 604, the dequantization unit 605, and the inverse orthogonal transformation unit 606. Inverse orthogonal transform unit 60
The output of 6 is output to the frame memory 6 in the adder 607.
10, the motion compensation unit 612, and the prediction signal of the previous frame supplied through the switch 613 are added, and the addition output is output as an output image signal from the terminal 609 via the digital-analog (D / A) conversion unit 608. . The output of the adder 607 is also input to the frame memory unit 610. The output of the frame memory unit 610 can be input to the motion compensation unit 612 via the switch 611, and can also be returned to the input side of the D / A conversion unit 608.

The error detecting section 602 detects a transmission error of the input coded data, and if there is an error, outputs an error detection signal to control the switch 611. At this time, the data of the block corresponding to the same pixel position of the previous frame stored in the frame memory 608 is reproduced as the output image signal.

The switch 613 is controlled by the intra-frame / inter-frame mode signal detected by the data separation unit 603. If the coded data has been subjected to intra-frame processing, the switch 613 is opened and subjected to inverse orthogonal transform. The added signal is directly output from the adder 607. However, when it is indicated that the mode signal is the inter-frame processed signal, the switch 613 is closed and the inter-frame predicted signal from the motion compensation unit 612 is added to the output of the inverse orthogonal transform unit 606. It has become.

In the above apparatus, the switch 611 and its control means apply the correction means when a transmission error occurs. As an example of using this correction method,
"JP-A-4-79689, image error correction method"
There is an example described in (Reference 1) or “JP-A-4-98987, Error Correction Method” (Reference 2). Here, the error correction method of Document 2 will be described with reference to the drawings.

According to this correction method, when data is divided in one field or one frame of a digital video signal, the data is divided into blocks, high efficiency coding is performed, and when recording / reproducing or transmitting, a block in which an error is detected is detected. , A reproduced image is obtained by replacing pixels in the same position in one field or one frame that are temporally adjacent to each other in the block.

FIG. 6 (A) is a diagram for explaining an error correction method in the case where data is divided into blocks within a frame and blocks in which an error is detected are replaced with blocks at the same position in adjacent fields. is there. In the figure, if an error is detected in the block 72a of the third field and the block 72b of the fourth field forming the second frame, the block 72a in the first half of the pixels forming this block is Replaced with block 71 at the same position in the field
The block 72b in one field in the latter half is replaced with the block 73 in the same position in the subsequent field.

According to such a method, in order to correct a block in which an error has occurred, the block can be replaced with a pixel located at the same position in a temporally adjacent field or frame, so that the deterioration of the image quality of the reproduced image can be prevented. You can make it less noticeable.

However, this method is effective to prevent the image quality deterioration from being noticeable when the image signal is a still image or a pattern with little movement, but when the image signal has a relatively large movement, On the contrary, there is a problem that the bankruptcy becomes conspicuous.

This state is shown in FIGS. 6 (B) and 6 (C). In order to make the problem easier to understand, replacement of a block in which an error has been detected is limited to a block at the same position one frame before.

FIG. 6B shows an error correction method when the movement of the person from the previous frame to the current frame is very small. If an error is detected in the block indicated by the dotted line in the current frame, it will be replaced with the block in the same position indicated by the dotted line in the previous frame, so the person in the reproduced image will be as shown by the shaded area in the figure, and distortion will occur. Is relatively uninteresting.

However, when the motion of the person is large, if the block at the same position in the previous frame is replaced, the person in the reproduced image becomes as shown by the diagonal lines in FIG. 6C, and the failure becomes very noticeable. . If the movement of the image (person) continues in the subsequent frame as well, even if it is replaced with a block at the same position in the subsequent frame (or field), there will be differences in appearance, and the distortion will be the same. It will be noticeable.

[0016]

As described above, the conventional error correction methods are effective in making the image quality deterioration inconspicuous when the video signal is a still image or a picture with little movement. However, when the video signal has a relatively large movement, there is a problem that the failure is conspicuous.

Therefore, according to the present invention, an error correction method that can obtain a reproduced image of good quality even when a transmission error occurs in the image digital transmission and a normal image signal is lost, and an image using the error correction method. An object of the present invention is to provide a transmitter / receiver for highly efficient coded signals.

[0018]

SUMMARY OF THE INVENTION According to the present invention, when a digital video signal is divided into a plurality of pixel blocks for high-efficiency coding and then transmitted, the transmitting side adds information including low-frequency component information of the video signal. The method comprises the step of sending out information and the coding coefficient information of the video signal to a transmission line, and at the receiving side,
When an error occurs in the coding coefficient information on the transmission line, an error correction method including a step of obtaining the reproduced video signal by decoding the pixel block by using the already transmitted additional information is provided. is there. Further, according to the present invention, when a digital video signal is divided into a plurality of pixel blocks and subjected to high-efficiency coding for transmission, additional information including low frequency component information of the video signal and coding of the video signal are transmitted on the transmitting side. Coefficient information and a transmission means for sending the coefficient information to the transmission path are provided, and at the receiving side, if an error occurs in the coded coefficient information on the transmission path, the pixel block is decoded using the already transmitted additional information. And has a means for obtaining a reproduced video signal,
The transmission means multiplexes a high-efficiency encoding means for dividing the input video signal into a plurality of pixel blocks for high-efficiency encoding, additional information indicating the characteristics of the pixel blocks, and a high-efficiency encoded output. And a means for sending a multiplexed signal to the transmission path.

[0019]

According to the above means, when an error occurs in the coding coefficient information on the transmission line, the intra-frame coded DC component containing a lot of important information as an image is moved using the motion vector of the block. It can be compensated and regenerated. That is, when the video signal has a relatively large motion, it works so as to improve the problem of the conventional error correction method that the failure is conspicuous in the block in which the error is detected.

[0020]

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

FIG. 1A shows an embodiment of the present invention.
The coded data introduced to the terminal 100 is temporarily stored in the buffer memory unit 101 and supplied to the error detection unit 102 and the data separation unit 103. The signal format of the encoded data is as shown in FIG. 1 (B), and the luminance (Y)
Block synchronization / address information, block mode information, intra-frame coded DC component information, and motion vector information are multiplexed as additional information before the coding coefficient of each of the signal and the C (color) signal.

In the data separation unit 103, the encoded data is separated into synchronization / address information, mode information, motion vector, intra-frame encoded DC component, and encoding coefficient. Sync / address information, mode information, motion vector,
The intra-frame coded DC component is the additional information here. The coding coefficients are the variable length decoding unit 104 and the inverse quantization unit 1.
05, and is supplied to the adder 107 via the inverse orthogonal transform unit 106. In the adder 107, the frame memory 110,
The prediction signal of the previous frame supplied from the motion compensation unit 112 is added, and the addition output is passed through the switch 108 and the digital-analog (D / A) conversion unit 104 to the terminal 115.
Is output as a reproduced image signal.

The switch 108 (operation will be described later) can selectively derive the output of the adder 107 or the motion compensation unit 112. The output of the switch 108 is output via the switch 109 (operation will be described later) to the frame memory unit 1.
Supplied to 10. Further, the output of the frame memory unit 110 is supplied to the input side of the motion compensation unit 112 or the D / A conversion unit 114 via the switch 111 (operation will be described later). The output of the motion compensation unit 112 is input to the adder 107 via the switch 113 (operation will be described later).

Here, each switch 108, 109, 11
The control signal generator of No. 3 will be clarified. The error detection unit 102 can obtain the coding coefficient error determination signal, and controls the switch 108 accordingly. Further, the error detection unit 102 can obtain the additional information error determination signal, and controls the switch 111 accordingly. The data separation unit 103 can extract the intra-frame / inter-frame mode signal, and thus the switches 113 and 10
Control 9 Furthermore, the data separation unit 103 can extract the motion vector, and thus the motion compensation of the motion compensation unit 112 can be performed.

The switch 108 initially selectively derives the output of the adder 107, but if there is an error in the coefficient due to the coding coefficient error determination signal supplied from the error detection unit 102, the motion compensation unit. It is switched to the DC component output end side of the frame 112.

Note that this system includes a variable length decoding unit 10
4, inverse quantizer 105, inverse orthogonal transformer 106, adder 1
07, the frame memory 110, the motion compensation unit 112, and the like, the decoding system also decodes the intra-frame coded DC component information of the additional information to obtain the intra-frame DC component.

When the additional information error detection signal indicates that the additional information has an error, the switch 111 supplies the output of the frame memory unit 110 to the input side of the D / A conversion unit 114 and there is no error. In the case of, the output of the frame memory unit 110 is supplied to the motion compensation unit 112 side.

The switch 113 is turned off (open) when the mode signal indicates intra-frame processing, and supplies the output of the motion compensation section 112 (inter-frame prediction signal output) to the adder 107 when it indicates inter-frame processing. To do.

Therefore, for the block in which the additional information has no error and the coding coefficient has an error, the switch 108 selects the intra-frame DC component from the motion compensating unit 112, so that the intra-frame important for the image is selected. A signal in which the DC component is motion-compensated can be reproduced. As a result, even if the video signal has a relatively large motion as shown in FIG. 2, a reproduced video signal can be obtained without conspicuous deterioration (an error in the coding coefficient is detected in block k in the figure). And). FIG. 3 shows a configuration example of the error detection unit 102 of FIG.

In the figure, the encoded data supplied to the terminal 200 is supplied to the data separation unit 201, and the block synchronization / address information, mode information, in-frame DC component,
The motion vector information and the coding coefficient are separated and supplied to the respective error detectors 202 to 206. Here, each error detection is performed, for example, based on whether the number of pieces of each information is equal to a prescribed number. If the pixel block is one orthogonal transform unit (assuming 8 × 8 pixels) and the motion vector is also detected in this unit, there are 64 coding coefficients and two motion vectors, one horizontal and one vertical. Good
If the pixel block is a collection of n orthogonal transform units, the transform coefficient is 64 × n and the motion vector is 2 × n.
The number of items and other information may be n. In this way 4
The presence or absence of an error in the kind of information is determined and input to the additional information error detector 207. If at least one of the four has an error, there is an error, and if all have no error, no error is supplied to the switch 111 from the terminal 208. In addition, the determination result of the coding coefficient is obtained from the terminal 209 in FIG.
Is supplied to the switch 108 as a coding coefficient error determination signal. FIG. 4 shows another embodiment of the present invention.

The signal format of the encoded data is shown in FIG.
1B is similar to the embodiment of FIG. Terminal 300
The encoded data received in step 1 is temporarily stored in the buffer memory 301 and supplied to the data separation section 302. In the data separation unit 302, the encoded data is separated into block mode information, motion vector information, intra-frame DC component, and coding coefficient, and the coding coefficient is the coefficient error detection unit 30.
3 and the switch 307.

The determination result of the coefficient error detection unit 303 is supplied to the switch 307 and the switch 317, and if it is determined that there is no transmission error in the coding coefficient, the switch 307.
Is closed, and the coding coefficient is decoded through the variable length decoding unit 311, the inverse quantization unit 312, and the inverse orthogonal transformation unit 313, and is supplied to the adder 316 through the switch 315. Adder 3
16, the signal from the switch 315 and the motion compensation unit 3
21 and the previous frame prediction signal supplied via the switch 322 are added, and the addition output is the D / A conversion unit 3
At 23, it is converted into an analog signal and is output to the output terminal 324 as a reproduction output image signal.

The additional information separated by the data separation unit 302 is supplied to the DC component error detection unit 304, the mode signal error detection unit 305, the motion vector error detection unit 306, and the switches 308, 309 and 310, respectively. These error detection units are determined by, for example, whether or not the number of pieces of each information is equal to a prescribed number. Each determination result is supplied to the additional information error detection unit 314. If all the determination results are error-free, an error determination signal is output from this detection unit 314 if there is no error or if there is at least one error. And is supplied as a control signal to the switch 315 and the switch 320.

The switch 308 is a DC component error detector 3
When the determination result of no transmission error is obtained in 04, the process is closed. At this time, the intra-frame DC component is supplied to the inverse quantization unit 312, and is supplied to the adder 316 via the inverse orthogonal transform unit 313. The switch 315 is controlled to close if there is no transmission error in the additional information and open if there is a transmission error.

The switch 317 is initially in a state in which the output of the adder 316 is selected, but is controlled so as to select the output from the motion compensation unit 321 if there is a transmission error in the coding coefficient. On the other hand, the switch 320 provided between the frame memory 319 and the motion compensating unit 321 gives the output of the frame memory 319 to the motion compensating unit 321 if there is no error in the additional information, and the frame 320 if the additional information has an error. It is controlled so that the output of the memory 319 is given to the D / A conversion unit 323.

The switch 322 supplies the output of the motion compensation circuit 321 to the adder 326, which is also controlled by the mode information and is controlled to be opened in the intraframe mode and closed in the interframe mode. .

According to the above circuit, if even one additional information has an error, the switch 320 causes the frame memory 319 to operate.
Is supplied to the input side of the D / A conversion unit 323, and the block data at the same position in the previous frame is directly interpolated and reproduced. If the additional information has no error and only the coding coefficient has a transmission error, the intra-frame DC component can be motion-compensated by the motion vector to be compensated and reproduced.

The present invention is not limited to the above-described embodiment. For example, the Y coding coefficient and the C coding coefficient are subjected to error detection separately, and a transmission error is detected only in the C coding coefficient. If detected, the Y coefficient may be subjected to normal processing, and only the C coefficient may be reproduced by performing motion compensation on the intra-frame DC component. The transmission order of the additional information is shown in FIG.
However, it goes without saying that other transmission orders may be used. Further, the present invention can be variously modified and implemented without departing from the gist thereof.

[0039]

As described above, according to the present invention,
When an error occurs in the coding coefficient information on the transmission path, the intra-frame coded DC component containing a lot of important information as an image can be reproduced by performing motion compensation using the motion vector of the block. Therefore, even if there is a relatively large motion in the video signal, failure does not stand out in the block in which the error is detected, so it is good even when a transmission error occurs in video digital transmission and normal image information is lost. It is possible to obtain a reproduced image of various quality.

[Brief description of drawings]

FIG. 1 is a circuit diagram showing an embodiment of the present invention and a diagram showing a video signal transmission block format according to the present invention.

FIG. 2 is an explanatory diagram shown for explaining the operation of the apparatus of the present invention.

FIG. 3 is a specific circuit diagram of the error detector of FIG.

FIG. 4 is a circuit diagram showing another embodiment of the present invention.

FIG. 5 is a circuit diagram showing a conventional decoding device.

FIG. 6 is a diagram for explaining a conventional correction method for correcting a transmission error block and its problems.

[Description of Codes] 101, 301 ... Buffer memory unit, 102 ... Error detection unit, 103, 302 ... Data separation unit, 104, 311 ...
Variable length decoding unit, 105, 312 ... Inverse quantization unit, 106,
313 ... Inverse orthogonal transform unit, 107, 316 ... Adder, 10
8, 109, 111, 113, 308 to 310, 31
5, 317, 320, 322 ... Switch, 110, 31
9 ... Frame memory unit, 115, 323 ... D / A conversion unit, 303, 304 ... DC component error detection unit, 305 ... Mode signal error detection unit, 306 ... Motion vector error detection unit, 314 ... Additional information error detection unit.

Claims (10)

[Claims] 1. When a digital video signal is divided into a plurality of pixel blocks for high-efficiency coding and then transmitted, additional information including low-frequency component information of the video signal and the coding of the video signal are transmitted on the transmitting side. Coefficient information and the step of sending the coefficient information to the transmission path, and at the receiving side, if an error occurs in the coded coefficient information on the transmission path, the pixel block is decoded using the already transmitted additional information. , An error correction method in high-efficiency video coding transmission, which comprises the step of obtaining a reproduced video signal. 2. The error correction in high-efficiency video coded transmission according to claim 1, wherein the transmission side sends the additional information to the transmission line prior to coding coefficient information of the video signal. Method. 3. The error in video high-efficiency coded transmission according to claim 1, wherein the transmission side sends the additional information to the transmission path next to the coding coefficient information of the video signal. Correction method. 4. The additional information is block synchronization / address information of a pixel block, intra-frame encoded DC component information of a video signal, motion vector information, or adaptive mode information, all or some of these. 2. The error correction method in video efficient coding transmission according to claim 1, wherein the additional information is combined and transmitted. 5. The decoding means, when there is no transmission error in the additional information and a transmission error in the coding coefficient information, performs the motion compensation on the intra-frame coded DC component information using the motion vector information. The error correction method in the video high-efficiency coded transmission according to claim 1, wherein the error correction method is a means for performing a reproduction video signal. 6. When the digital video signal is divided into a plurality of pixel blocks for high-efficiency coding and then transmitted, additional information including low-frequency component information of the video signal and the coding of the video signal are transmitted on the transmitting side. Coefficient information and a transmission means for sending the coefficient information to the transmission path are provided. On the receiving side, when an error occurs in the coded coefficient information on the transmission path, the pixel block is decoded using the already transmitted additional information. And a means for obtaining a reproduced video signal, wherein the transmission means divides the input video signal into a plurality of pixel blocks to perform high-efficiency encoding, and a characteristic of the pixel blocks. A transmitter / receiver in high-efficiency coded transmission, comprising: means for multiplexing the additional information shown and a high-efficiency coded output and sending a multiplexed signal to the transmission path. 7. The additional information indicating the characteristics of the pixel block is block synchronization / address information of the pixel block, intra-frame coded DC component information of the video signal, motion vector information, or adaptive mode information. 7. The transmitter / receiver in high efficiency coded transmission according to claim 6, further comprising means for transmitting all or some of the information as additional information. 8. The high-efficiency code according to claim 6, wherein the means for multiplexing the additional information and the high-efficiency coded output sends the additional information to the transmission line at the head of the transmission image block data. Transmitter / receiver for digitalized transmission. 9. The means for obtaining the reproduced video signal comprises means for separating the received multiplexed signal into additional information and a high-efficiency coded output, decoding means for decoding each separated output, and each decoding means. 7. A transmitter / receiver in high-efficiency coded transmission according to claim 6, further comprising means for obtaining a reproduced video signal by using an output. 10. The means for obtaining the reproduced video signal, when there is no transmission error in the additional information and a transmission error occurs in the coding coefficient information, the intra-frame coded DC component information included in the additional information. 7. The transmitter / receiver in high-efficiency coded transmission according to claim 6, wherein motion compensation is performed using motion vector information to obtain the reproduced video signal.