JP2001103479A - Dynamic image decoder - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress the occurrence of a frozen image and a blackout that are unnatural images, when the resolution of a displayed image is changed. SOLUTION: Coded data are decoded simultaneously, by using a high resolution coding data decoder and a low resolution coding data decoder. The result of the high resolution decoding is used for data, where no error takes place and displayed and the result of the low-resolution decoding is converted into data with high-resolution, when an error exists in the result of the low resolution decoding and the converted data are displayed. When the error rate reaches a high rate, the data as the result of the low-resolution decoding only are used and displayed. In the case that the scanning frequency of a display section can be changed, the resolution is not changed, and the data are displayed by adopting a lower scanning frequency. One and same circuit and a common memory are used in time division, to reproduce the data as the result of the high resolution decoding and the low-resolution decoding.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a moving picture decoding apparatus for reducing the influence of a decoding error on a display picture when decoding an encoded moving picture.

[0002]

2. Description of the Related Art Conventionally, an analog system for performing analog modulation on a moving image signal such as an NTSC system or a PAL system has been used for transmitting a moving image. On the other hand, in recent years, there has been an increasing demand for handling moving images digitally.However, since the amount of data when a moving image is digitized becomes enormous, the moving image that has been digitized in the same transmission band as the analog method is used. Cannot be transmitted, and the capacity required for the recording medium becomes enormous. Therefore, the amount of data to be transmitted has been reduced by using a digital moving image encoding technology represented by the MPEG system.

[0003] When wireless is used for transmitting moving images, the effect of noise changes the transmission characteristics every moment. This may be because the transmitting side or the receiving side is moving, or because the amount of attenuation of radio waves changes due to rainfall. Also, in storage media, due to aging of media and storage conditions,
The magnitude of the influence of noise when reading is changed.

FIG. 7 is a conceptual diagram showing the quality of analog and digital decoded images when the transmission path is affected by noise. In the analog system, even when the noise level is low, the influence of noise is exerted on the decoded image. However, even when the noise level is high, the content of the decoded image can be determined to some extent. On the other hand, in the digital system, when the noise level is low, the error is corrected by the error correction code and the noise is not affected, but when the noise level is high, the error cannot be corrected completely, and the quality of the decoded image is reduced. . By using a code with improved error correction capability, noise resistance can be increased. However, such a code has a large degree of redundancy, so that compression efficiency is reduced.

[0005] Therefore, when the transmission is performed in a digital system,
Transmission of the same content using multiple error correction codes with different error correction capabilities, and transmission of the content whose power rate is concentrated to content whose error rate is to be reduced, especially in the field of broadcasting Is being done. As described above, since the compression efficiency decreases when a code with a high error correction capability is used, when using a code with a high error correction capability, moving image coding is performed on the content expressed in low resolution.
When a code having a low error correction capability is used, the content expressed in a high resolution is encoded with a moving image. In general, a high-resolution signal format is 1920 horizontal pixels × 1080 vertical pixels × frame rate of 30 Hz or 720 horizontal pixels × 4 vertical pixels.
The signal is transmitted at 80 pixels × frame rate of 30 Hz, and the low-resolution signal format uses horizontal 352 pixels × vertical 240 pixels × frame rate 30 Hz or horizontal 176 pixels × vertical 120 pixels × frame rate 30 Hz.

On the receiving side, the frequency of transmission errors of high resolution and low resolution is constantly monitored. If the frequency of transmission errors is low, decoding and display are performed using coded data of high resolution, and the frequency of transmission errors is high. In this case, decoding and display are performed using low-resolution encoded data. Using the error rate detected by the error correction unit or the error rate detected by the video decoding unit, etc., a resolution switching operation for decoding and displaying is performed, thereby achieving the analog system as shown by the broken line in FIG. Thus, it is possible to obtain a decoded image with little influence of noise. The low-resolution reproduced image is displayed after performing sampling rate conversion in order to match the scanning frequency of the display unit.

[0007]

However, if the threshold value of the error rate at the time of performing the switching operation is low, although most areas can be decoded and displayed with high-resolution data, the area can be decoded at low resolution. Decoding and reproduction will be performed, and if the setting of the threshold is high, decoding and display will be performed at a high resolution even though most of the decoded images are broken due to errors. In addition, when the moving picture coding method using the motion compensation is used, as shown in FIG. 8, since the reference picture for the motion compensation is not created at the switching point from the high resolution to the low resolution, the normal decoding is performed. Cannot be performed, and the screen freezes or blacks out temporarily.

In order to prevent this, it is necessary to always perform high-resolution and low-resolution decoding. However, an existing single-stream video decoding device needs to have a double decoding unit. There is a concern that the circuit scale will increase. Also,
A multi-stream video decoding device aims to decode moving images of different contents having the same resolution.If one of the resolutions is low, the processing capacity remains, and an error occurs on the high resolution side. Error correction could not be performed from the decoding result on the low resolution side, which is being decoded at the same time when the error occurred.

Further, when a low-resolution moving image is converted into a high-resolution image, distortion generated by moving image encoding is expanded in the spatial direction by a resolution conversion filter, so that an unnatural reproduced image is obtained.

The present invention has been made to solve the above-described problem. First, it is an object of the present invention to perform display switching stepwise due to a change in error rate, and secondly, to switch to display of low-resolution data. Thirdly, it is intended to reduce unnaturalness of a displayed image at the time, and thirdly, to suppress an increase in circuit scale when the present method is used.

[0011]

According to a first aspect of the present invention, low-resolution coded data and high-resolution coded data are simultaneously coded at low resolution and high resolution. If both data are decoded at the same time and a decoding error occurs during decoding of the high resolution coded data, and no error occurs in the low resolution decoding at the same position on the screen, the effect of the error is high resolution. Is replaced by correcting the decoded image of low resolution with the decoded image of.

According to the second aspect of the present invention, when the frequency of high-resolution decoding errors is high, display is performed using only low-resolution decoded images, and when the frequency of high-resolution decoding errors is low, high-resolution decoding images are displayed. The decoded image having a lower resolution is corrected and replaced with the decoded image, and the decoded image is displayed.

In contrast to the case where motion compensation is used for video coding, according to the third aspect of the present invention, a picture on which a decoding error has occurred during decoding of high-resolution coded data is referred to in subsequent decoding. When used as an image, a decoded image with a high resolution that is affected by an error is replaced with a decoded image with a low resolution, and is used as a reference image for subsequent decoding.

In the fourth aspect of the present invention, when the display is performed using only the low resolution decoded image, the display is performed by switching the scanning frequency of the display unit to the low resolution.

According to the fifth aspect of the present invention, error decoding is performed by performing time-division processing on a high-resolution coded data decoding unit and a low-resolution coded data decoding unit using the same circuit and a common memory. In order to improve the robustness, an increase in the circuit scale when decoding is always duplicated is suppressed.

[0016]

FIG. 1 shows a first embodiment of the present invention. Encoded data 101 is encoded data separating section 102
, The high-resolution encoded data 103 is output to the high-resolution encoded data decoding unit 105, and the low-resolution encoded data 104 is output to the low-resolution encoded data decoding unit 106. The high-resolution coded data decoding unit 105 decodes the high-resolution coded data 103 and outputs a high-resolution decoding result 107 and high-resolution decoding error information 108 to the error correction unit 111. The low-resolution coded data decoding unit 106 decodes the low-resolution coded data 104,
Low-resolution decoding result 109 and low-resolution decoding error information 11
0 is output to the error correction unit 111. Error correction unit 11
In 1, when the high-resolution decoding error information 108 is valid, the low-resolution decoding error information 110 in the corresponding area and its surroundings is referred to, and when the decoding error information is invalid, the low-resolution decoding result 109 is corrected to high. Decoded image 1 of resolution
12 is output.

FIG. 2 shows an example of the reconstruction by the error correction unit 111.
Show. The high-resolution decoding result 201 includes a pixel 202 decoded without an error and an area 2 that could not be decoded due to an error.
03 exists. All the low-resolution decoding results 204 are decoded without error, and there are pixels 205 not used for high-resolution error correction and pixels 206 used for error correction. The error correction unit 111 performs resolution conversion on the pixel 206 used for error correction, and creates a high-resolution corrected pixel 207. The high-resolution correction pixel 207 is attached to the area 203 of the high-resolution decoding result 201 that could not be decoded due to an error.

In the resolution conversion processing, a linear filter may be used, or a non-linear filter for detecting an edge of an image and performing non-linear processing may be used. Also, in order to reduce the unnaturalness of the boundary between the pixel 202 decoded without a high-resolution error and the correction pixel 207 created from the low-resolution decoding result, a filter may be applied only to the boundary.

FIG. 3 shows a second embodiment of the present invention. The coded data 301 is input to the coded data separation unit 302, and the coded data 303 of high resolution is transmitted to the coded data decoding unit 305 of high resolution.
4 is output separately to the low-resolution coded data decoding unit 306. The high-resolution coded data decoding unit 305 decodes the high-resolution coded data 303 and outputs a high-resolution decoded result 3
07 to the error correction unit 311 and the high resolution decoding error information 308 to the error correction unit 311 and the error rate determination unit 313. Low-resolution encoded data decoding unit 306
Then, the low-resolution coded data 304 is decoded, the low-resolution decoding result 309 is output to the error correction unit 311, and the low-resolution decoding error information 310 is output to the error correction unit 311 and the error rate determination unit 313. The error rate determination unit 313 obtains each error rate from the high resolution decoding error information 308 and the low resolution decoding error information 310,
When the high-resolution error rate exceeds the threshold, the low-resolution use signal 314 is output to the error correction unit 311. When the low-resolution use signal 314 is invalid, and when the high-resolution decoding error information 308 is valid, the error correction unit 311 refers to the low-resolution decoding error information 310 in the relevant area and its surroundings to obtain the low-resolution decoding error information. Is invalid, the low-resolution decoding result 309 is corrected and a high-resolution decoded image 312 is output. When the low-resolution use signal 314 is valid, all the low-resolution decoding results 309 are corrected and a high-resolution decoded image 312 is output.

FIG. 4 shows a third embodiment of the present invention. The coded data 401 is input to the coded data separation unit 402, and the high-resolution coded data 403 is sent to the high-resolution coded data decoding unit 405 and the low-resolution coded data 40
4 is output to the low-resolution encoded data decoding unit 406 separately. The high-resolution encoded data decoding unit 405 decodes the high-resolution encoded data 403 using the high-resolution reference image data 413 output from the high-resolution decoding reference image memory 415, and outputs the high-resolution decoded result 407 and the high-resolution decoded result. The decoding error information 408 is output to the error correction unit 411. The low-resolution coded data decoding unit 406 decodes the low-resolution coded data 404 using the low-resolution reference image data 414 output from the low-resolution decoding reference image memory 416, and outputs a low-resolution decoded result 409 and a low-resolution decoded result. The decoding error information 410 is output to the error correction unit 411.
Low-resolution decoding result 409 used as the next reference image
Are stored in the low-resolution decoding reference image memory 416. When the high-resolution decoding error information 408 is valid, the error correction unit 411 refers to the low-resolution decoding error information 410 in the corresponding area and its surroundings, and corrects the low-resolution decoding result 409 when the decoding error information is invalid. To output a high-resolution decoded image 412. The high-resolution decoded image 412 used as the next reference image is stored in the high-resolution decoding reference image memory 415.

As another operation of this embodiment, when an error occurs, the high-resolution coded data decoding unit 405 reads the high-resolution reference image data 413 at the same position as the error occurrence area from the high-resolution decoding reference image memory 415. When the error concealment is performed in advance by performing the error concealment and the high-resolution decoding result 407 is output to the error correction unit 411, the high-resolution decoding result 407 and the high-resolution decoding error information 408 are output without performing the error concealment.
Can be switched in accordance with the frequency of the error, the size of the area, or the amount of movement of the error-free peripheral area.

FIG. 5 shows a fourth embodiment of the present invention. The encoded data 501 is input to the encoded data separation unit 502, and the high resolution encoded data 503 is sent to the high resolution encoded data decoding unit 505, and the low resolution encoded data 50
4 is output separately to the low-resolution coded data decoding unit 506. The high-resolution coded data decoding unit 505 decodes the high-resolution coded data 503 and outputs a high-resolution decoded result 5.
07 to the error correction unit 511 and the high-resolution decoding error information 508 to the error correction unit 511 and the error rate determination unit 513. Low-resolution encoded data decoding unit 506
Then, the low-resolution encoded data 504 is decoded, the low-resolution decoding result 509 is output to the error correction unit 511, and the low-resolution decoding error information 510 is output to the error correction unit 511 and the error rate determination unit 513. The error rate determination unit 513 obtains each error rate from the high resolution decoding error information 508 and the low resolution decoding error information 510,
When the high-resolution error rate exceeds the threshold, the low-resolution use signal 514 is sent to the error correction unit 511 and the display unit 5.
15 is output. When the low resolution use signal 514 is invalid, the error correction unit 511 outputs the high resolution decoding error information 5
If 08 is valid, the low-resolution decoding error information 510 in the corresponding area and its surroundings is referred to. If the low-resolution decoding error information is invalid, the low-resolution decoding result 509 is corrected and the high-resolution decoded image 512 is corrected. Is output to the display unit 515. When the low resolution use signal 514 is valid, all the low resolution decoding results 509 are output to the display unit 515 as decoded images 512 as they are. The display unit 515 changes the scanning frequency according to the low resolution use signal 514 and performs display.

FIG. 6 shows a fifth embodiment of the present invention. Encoded data separation section 602, encoded data decoding section 611,
An error correction unit 614 and a memory unit 604 are provided. The memory unit 604 logically includes a high-resolution coded data memory 620, a low-resolution coded data memory 621, a high-resolution reference image memory 622, a low-resolution reference image memory 623, a high-resolution error correction memory 624, and a low-resolution error correction. Memory 625 for use. In addition, the encoded data decoding unit 611 repeats decoding of high-resolution data and decoding of low-resolution data alternately. Hereinafter, a specific operation will be described.

The coded data 601 is input to a coded data separation unit 602, where the coded data 601 is separated into high-resolution coded data and low-resolution coded data. The low-resolution coded data 607 is written to the low-resolution coded data memory 621 via the memory bus 605. The encoded data decoding unit 611 alternately decodes high-resolution data and low-resolution data. When decoding the high-resolution data, the high-resolution coded data 608 is read from the high-resolution coded data memory 620, and if necessary, the high-resolution reference image 609 is read from the high-resolution reference image memory 622 and decoded. 613, error information 612
And the decoding resolution information 603 is output to the error correction unit. When decoding the low-resolution data, the low-resolution coded data 610 is read from the low-resolution coded data memory 621, the low-resolution reference image 626 is read from the low-resolution reference image memory 623 as necessary, and the decoding is performed. 613, error information 612 and decoding resolution information 60
3 is output to the error correction unit, and if necessary, the decoding result 613 is output.
Is written in the low-resolution reference image memory 623. The error correction unit 614 refers to the decoding resolution information 603, and outputs the decoding result 613 to the high-resolution error correction memory 624 as the decoding result 613 when high-resolution decoding is being performed. In this case, the decoding result 613 is output to the low-resolution error correction memory 625, and the error information 612 is stored therein. If an error has occurred in a part of the high-resolution decoding result after both decoding results have been completed, the high-resolution error correction data 616 is read from the high-resolution error correction memory 624 for an error-free part. Play image 61
9 is output, and the portion where an error has occurred is stored in the low-resolution error correction memory 625 from the low-resolution error correction data 618.
, And after performing resolution conversion, a reproduced image 619 is output. Also, when used later as a reference image,
The reproduced image 619 is written in the high-resolution reference image memory 622.

[0025]

As described above, according to the present invention, moving images having the same contents are encoded at a low resolution and a high resolution, and those recorded in advance are simultaneously reproduced or encoded data transmitted simultaneously is decoded. When an error is detected in a partial area when decoding high-resolution data, the error correction can be performed using the low-resolution data that has been simultaneously decoded in the error area. become. Even if the error rate changes over time and the frequency of errors suddenly worsens, it is possible to switch the display continuously using low-resolution data that is being played back at the same time, and freeze or blackout the screen. Can be prevented.

Furthermore, by lowering the scanning frequency of the display unit when only low-resolution data is displayed, it is possible to prevent an increase in coding noise due to an interpolation filter for resolution conversion.

Further, by slightly improving the performance of the decoding unit, it becomes possible to share the decoding unit for the high-resolution data and the decoding unit for the low-resolution data, thereby increasing the circuit scale of the moving picture decoding apparatus with enhanced error tolerance. Can be suppressed.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a first embodiment of the present invention.

FIG. 2 is a view for explaining an outline of the operation of an error correction unit according to the first embodiment of the present invention.

FIG. 3 is a block diagram showing a configuration of a second embodiment of the present invention.

FIG. 4 is a block diagram showing a configuration of a third embodiment of the present invention.

FIG. 5 is a block diagram showing a configuration of a fourth embodiment of the present invention.

FIG. 6 is a block diagram showing a configuration of a fifth embodiment of the present invention.

FIG. 7 is a conceptual diagram illustrating resistance to errors in a conventional analog system and digital system.

FIG. 8 is a diagram illustrating a period in which display at a resolution switching point is not performed in a conventional example.

[Explanation of symbols]

101, 301, 401, 501, 601 Encoded data 102, 302, 402, 502, 602 Encoded data separation units 103, 303, 403, 503 High-resolution encoded data 104, 304, 404, 504 Low-resolution code Coded data 105, 305, 405, 505 high-resolution coded data decoding unit 106, 306, 406, 506 low-resolution coded data decoding unit 107, 307, 407, 507 high-resolution decoding result 108, 308, 408, 508 high-resolution Decoding error information 109, 309, 409, 509 Low-resolution decoding result 110, 310, 410, 510 Low-resolution decoding error information 111, 311, 411, 511, 614 Error correction unit 112, 312, 412, 512 Decoded image 313, 513 Error rate determination unit 314 514 Low-resolution use signal 413 High-resolution reference image data 414 Low-resolution reference image data 415 High-resolution decoding reference image memory 416 Low-resolution decoding reference image memory 515 Display unit 603 Decoding resolution information 604 Memory unit 605 Memory bus 606 High resolution Encoded data 607 Low-resolution encoded data 608 High-resolution encoded data 609 High-resolution reference image 610 Low-resolution encoded data 611 Encoded data decoding unit 612 Error information 613 Decoding result 615 Data before high-resolution correction 616 For high-resolution correction Data 617 Data before low-resolution correction 618 Data for low-resolution correction 619 Playback image 620 High-resolution encoded data memory 621 Low-resolution encoded data memory 622 High-resolution reference image memory 623 Low-resolution reference image memo Re 624 High resolution error correction memory 625 Low resolution error correction memory 626 Low resolution reference image

Claims (5)

[Claims] 1. A moving image decoding apparatus and method for encoding moving images of the same content at low and high resolutions and reproducing previously recorded ones simultaneously or decoding encoded data transmitted at the same time. It has a separation unit that separates low-resolution encoded data and high-resolution encoded data, decodes both low-resolution encoded data and high-resolution encoded data simultaneously, and decodes high-resolution encoded data. If a decoding error occurs during processing and no error occurs in decoding at a low resolution at the same position on the screen, the decoded image at a higher resolution that is affected by the error is replaced with a decoded image at a lower resolution and replaced. A moving image decoding apparatus. 2. The video decoding apparatus and method according to claim 1, wherein when the frequency of the high-resolution decoding error is high, the display is performed using only the low-resolution decoded image, and the frequency of the high-resolution decoding error is reduced. A moving picture decoding apparatus characterized in that when the picture is low, a decoded picture with a high resolution is replaced with a decoded picture with a low resolution and displayed. 3. A moving picture decoding apparatus and method according to claim 1 and 2, wherein a decoding error occurs during decoding of high resolution coded data. If the screen is used as a reference image for subsequent decoding, the high-resolution decoded image affected by the error is replaced by correcting the low-resolution decoded image,
A moving picture decoding apparatus characterized by being used as a reference picture for subsequent decoding. 4. When the moving image decoding apparatus according to claim 2 performs display using only a low-resolution decoded image, it is preferable to switch the scanning frequency of the display unit to a low-resolution image and perform display. A moving picture decoding apparatus characterized by the following. 5. The moving picture decoding apparatus according to claim 1, wherein said decoding unit decodes high-resolution coded data and decodes low-resolution coded data. A moving picture decoding apparatus characterized in that the sections are time-divisionally processed using the same circuit and a common memory.