JPH08317386A - Decoding method and apparatus for interlaced coded image - Google Patents

Abstract

(57) [Abstract] [Purpose] A decoding method and apparatus for an interlaced image for decoding an interless image encoded by an international standardization method such as MPEG, and recovers without display deterioration even when a decoding error occurs. It is an object of the present invention to provide a decoding method and apparatus for interlaced coded images that can be performed by When an error is detected in the coded data when the variable length decoding unit 12 decodes the coded data in which the interlaced image is coded in units of fields forming a frame on the coding device side, an error recovery is performed. An error notification is sent to the control unit 17, and the error recovery control unit 17 determines whether the data of the field in which an error is detected according to the error notification from the variable length decoding unit 12 is the first field, the second field, or Next, the data of the decoded field adjacent to the error detection field is used as a substitute depending on whether the decoded field is the first field or the second field.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and apparatus for decoding an interless coded image, and more particularly to an interless coded image for decoding an interless image coded by an international standardization system such as MPEG. Decoding method and device.

To realize digital transmission of moving images and storage services, it is essential to efficiently compress the amount of information. Therefore, various methods have been studied as a moving picture coding method. Among them, interframe prediction is a method for removing the correlation in the time axis direction. In this method, when the first frame is coded, all the information of this frame is coded, but the subsequent frames are the differences from the previously coded frame. This method is particularly effective when the image is almost stationary or when only a part of the screen is moving.

From the above, interframe prediction is
The H.264 standard, which is an international standardization method for moving image coding, 261 MP
It is used in EG-1 and MPEG-2. MPE
G-2 is an extension of MPEG-1 and has a system capable of efficiently encoding an interlaced image such as an NTSC signal or a PAL signal, and can be encoded as an independent image for each field. It is possible.

When the data encoded by such an encoding method is to be decoded, the data of the frame previously encoded is required. Therefore, if the decoding cannot be performed due to an error, etc. No image can be obtained. Therefore, it is necessary to compensate the data even when an error or the like occurs so that the next image can be decoded.

[0005]

2. Description of the Related Art FIG. 6 shows a block diagram of an interlaced image. In the interlaced image, one frame 20 is composed of two fields 21 and 22. The frame 20 is configured by alternately combining two fields 21 and 22 for each line, and a field 21 located above is called a top field 21 and a field 22 located below is called a bottom field 22. There is. Of the two fields 21 and 22, the field displayed first in time is called the first field, and the field displayed later is called the second field.

Further, either the top field or the bottom field may be the first field, but once the one is determined, the order of the fields does not change thereafter. Below, in order to simplify the description, the top field will be described as the first field. The same applies to the opposite case.

By the way, in the case of performing image transmission using interframe coding, if data error occurs in the communication channel, the decoded image is greatly deteriorated. Therefore, if an image cannot be played back due to a decoding error in a certain frame, all coded frames that use this frame as a reference frame cannot be decoded, and display is continued until the next frame without inter-frame prediction is transmitted. It will not be possible.

Therefore, when an error occurs, the image that can be decoded immediately before the decoding error is repeatedly displayed (freeze state) until the next image can be decoded.

[0009]

However, in the conventional decoding method for interlaced coded images, if a decoding error occurs during decoding, the image that can be decoded immediately before the decoding error is repeatedly displayed, so that a moving image or the like is displayed. In this case, there is a problem in that the image is stopped until decoding becomes possible and the viewer feels uncomfortable.

The present invention has been made in view of the above points, and provides a decoding method and apparatus for an interlaced coded image capable of displaying an image with less discomfort by performing decoding as soon as possible when a decoding error occurs. The purpose is to

[0011]

FIG. 1 shows the principle of the present invention. The decoding method of an interlaced coded image according to the present invention is a method of decoding coded data generated by a coding method in which an interless image is coded in units of fields constituting a frame.

In the method for decoding an interlaced coded image according to the present invention, the error detection procedure 1 detects an error in the data of the decoded field. Further, in the data substitution procedure 2, when an error is detected in the error detection procedure 1, the data of the field which is close to the field in which the error is detected and which is decoded is substituted for the data in which the error is detected.

According to a second aspect of the present invention, the frame is composed of a first field and a second field, and the data substituting procedure decodes before the second field if an error is detected in the error detecting procedure. First done
First substituting field data for the second field
It is characterized by having a use procedure of.

According to a third aspect of the present invention, when the data substitution procedure is the second field where the error is detected in the error detection procedure and the second field is the field decoded after the error is detected, Second decrypted
It is characterized by having a second substitution procedure for substituting the data of the field for the data of the first field.

According to a fourth aspect of the present invention, in the data substitution procedure, a field in which an error is detected in the error detection procedure is a first field, and in a field decoded after the error is detected, an error is detected in the first field. In the case of the second field forming a field and a frame, it is characterized by having a third substitution procedure for substituting the decoded second field data for the first field data.

According to a fifth aspect of the present invention, in the data substitution procedure, the field in which an error is detected in the error detection procedure is the first field, the field decoded after the error is the second field, and the error is detected. Is not formed in the same frame as the detected first field, the decoded second field data is decoded into the second field.
It is characterized by having a fourth substitution procedure for substituting the data of the first field forming the same frame as the field.

According to a sixth aspect of the present invention, the data substitution procedure is substituted by the fourth substitution procedure and the data of the formed frame is substituted for the data of the frame including the first field in which the error is detected. It is characterized by having a procedure.

[0018]

According to the first aspect of the present invention, when a decoding error is detected in the error detecting procedure, it is possible to decode the frame by substituting the data of the adjacent decoded field. On the other hand, a quick frame restoration operation is possible, frame loss of an image due to an error can be minimized, and decoding can be continuously performed without affecting subsequent decoding.

According to the second aspect, in the case where one frame is composed of the first field and the second field, when the field in which the error is detected is the second field, the first field decoded before the second field. The frame of the field in which the error has occurred is restored by substituting the data of.

According to the third aspect, when the field in which the error is detected is the second field and the field decoded after the error is the second field, the error is detected in the second field decoded. The frame is restored by substituting the generated first field.

According to claim 4, when the field in which the error is detected is the second field which forms a frame with the first field after the error is detected in the first field, the decoding is performed. Secondized
The frame is restored by substituting the field for the first field in which an error was detected.

According to a fifth aspect of the present invention, the field in which an error is detected is the first field, the field which is decoded after the error is detected is the second field, and the same frame as the first field in which the error is detected. When the frame is not formed, the frame is restored by substituting the decoded second field for the first field forming the same frame as the decoded second field.

According to the sixth aspect, the lost frame is restored by substituting the frame restored in the fifth aspect for the frame of the field in which the error is detected. According to the seventh aspect, when a decoding error is detected by the error detecting means, the data of the adjacent decoded field is substituted for the field in which the error is detected to restore the frame. It is possible to quickly and quickly perform frame restoration operation, and it is possible to minimize the loss of image frames due to errors.
Further, since the frame is restored, there is no influence on the subsequent decoding.

[0024]

1 is a block diagram of an embodiment of the present invention. This embodiment is based on the international standardization method H.264. 261, M
A decoding device for decoding a moving image encoded by a moving image encoding method such as PEG1 and MPEG2 will be described.

The decoding device 11 of the present embodiment includes a variable length decoding unit 12 for restoring encoded data to original data, and an inverse quantization unit for performing inverse quantization processing on the data restored by the variable length decoding unit 12. 13, an inverse discrete cosine transform unit 18 that performs an inverse discrete cosine transform process on the dequantized data, a frame memory 14 that holds a decoded image, a data and a frame memory 14 that are inverse discrete cosine transformed by the inverse discrete cosine transform unit 18. Adder 15 for obtaining a decoded image by inter-frame prediction and the frame memory 1
When an error occurs during data decoding in the display unit 16 and the variable length decoding unit 12 that display the decoded image held in 4, the data in which the error has occurred is compensated and the influence on the image is reduced. The error recovery control unit 17 for controlling

The variable-length decoding unit 12 uses the H.264 standard, which is an international standardized system, on the encoding device side. 261, MPEG1, MPEG
The encoded data is supplied to the moving image encoded by the moving image encoding method such as 2. The variable length decoding unit 12 decodes the encoded data and notifies the error recovery control unit 17 of an error when a data error occurs.

The decoded data decoded by the variable length decoding unit 12 is supplied to the inverse quantization unit 13. Inverse quantizer 13
Then, the inverse quantization processing is performed on the decoded data output from the variable length decoding unit 12, and the result is supplied to the inverse discrete cosine transform unit 18. The inverse quantized data is supplied from the inverse quantization unit 13 to the inverse discrete cosine transform unit 18. The inverse discrete cosine transform unit 18 performs an inverse discrete cosine transform process on the inverse quantized data output from the inverse quantum unit 13, and supplies the result to the adder 15.

The adder 15 includes an inverse discrete cosine transform unit 18
The inverse discrete cosine transform processing result is supplied from
Reference image data used for inter-frame prediction is supplied from the frame memory 14, and decoding is performed by adding the inverse discrete cosine transform processing result from the inverse discrete cosine transform unit 18 and the reference image data from the frame memory 14. Generate image data. The decoded image data decoded by the adder 15 is supplied to the frame memory 14. The frame memory 14 holds the decoded image data supplied from the adder 15. The decoded image data stored in the frame memory 14 is supplied to the display unit 16. The display unit 16 outputs an image according to the image data supplied from the frame memory 14.

When the error recovery controller 17 receives an error notification from the variable length decoder 12, the error recovery controller 17 substitutes the image data of another field held in the frame memory 14 for the field in which the error occurred. Then, the data in the frame memory 14 is rewritten.

FIG. 3 is a flowchart of the data compensation operation when an error occurs by the error recovery controller 17 according to one embodiment of the present invention, and FIGS. 4 and 5 show the data compensation operation when an error occurs by the error recovery controller 17. An explanatory view is shown. FIG.
As shown in (A), the interlaced image is a frame FL.
Is composed of two fields TF and BF. The frame FL is configured by alternately combining two fields TF and BF for each line, and a field located above is called a top field TF and a field located below is called a bottom field BF. Also, of the two fields, the field that is displayed earlier in time is the first field, and the field that is displayed later is the second field.
Call it the field.

When the error recovery control unit 17 detects an error in the data, it determines whether the data in which the error is detected corresponds to the image data of the first field or the image data of the second field ( Step S1
_-1 , S1 _-2 ). When it is determined in step _S1-2 that an error has been detected in the image data corresponding to the second field, the error recovery control unit 17 then determines whether an error has occurred before the field BF1 as shown in FIG. 4B. the decoded image data a of the first field FB1 error decoded substitute the decoded image data of the second field BT1 that occurred (step S1 _-3).

By the above, the correction of the field in which the error occurred
You can make a payment. Error recovery that can obtain 1 frame of data
The return controller 17 decodes when the next field is decoded.
The converted field is the first field or the second field
It is determined (step S1 _-Four, S1_-Five). Step S
1_-FiveIf the field decoded by is the first field
Ends a series of data compensation processing.

Further, data of the second field BF2 then decoded field in step S1 _-5 is decoded as in the case of the second field shown in FIG. 4 (C) B
Is substituted for the data B of the first field TF2 forming a pair with the second field BF2 (step _S1-6 ).

As described above, the fields that could not be decoded are compensated and the image data for one frame is restored. Therefore, images can be continuously displayed without affecting the image prediction of the data that follows. If the field in which the error is detected is the first field in step _S1-2 , the error recovery control unit 17 waits for the next field to be decoded, and the next decoded field becomes the first field. It is determined whether it is the 1st field or the 2nd field (steps S1 _-7 and S1 _-8 ).

In step _S1-8 , when the decoded field is the first field, it does not affect the subsequent frame forming processing, so that the processing ends. If the field decoded in step _S1-8 is the second field, it is determined whether the second field decoded next is the field showing the same frame as the first field in which an error is detected. (Step S
_1-9 ).

In step _S1-9 , if the decoded second field is a field forming the same frame as the first field in which an error is detected, the error recovery control unit 17 shows in FIG. 5 (A). Thus, the data C of the second field BF3 which is paired with the first field TF3 in which the error is detected is substituted for the data of the first field TB3 in which the error is detected (step S).
1 _-10 ).

As described above, the frame including the field in which the error is detected can be restored, and the subsequent processing is not affected. In step _S1-9 , if the next decoded field is the second field forming a frame different from the first field in which the error is detected, the error recovery control unit 17 determines that the error recovery control unit 17 shown in FIG. ), The decoded data D of the second field BF4 is paired with the second field BF4, and is substituted for the undecoded data of the first field TF4 to generate the frame FL2 (step S1 _-11 ). . Next, the error recovery control unit 17
Is the second field B restored as shown in FIG.
An error is detected in the frame FL2 generated by substituting the data D of F4 for the first field TF4, and the frame F supposed to be generated by the first field TF3 and the second field BT3 forming a pair with the first field TF3.
Substitute for L1 (step _S1-12 ).

By the above, the frame FL2 is restored,
The subsequent processing is not affected, and the frame FL1 that should not be restored can be restored and the continuity of the frame can be maintained. As described above, according to the present embodiment, the two fields constituting the frame are sampled at a very short time interval, so that the images are almost the same. Therefore, by encoding the frame for each field, even if one of the fields has a decoding error, the other field can be decoded.By copying the decoded image, the original encoded image can be decoded correctly. It is possible to obtain an image that is almost similar to the image in the case. As a result, the subsequent frame using inter-frame prediction can also be decoded, and the processed image can be displayed smoothly.

Therefore, when an interlaced image is encoded in field units, even if a decoding error occurs, the frame whose frame is the predicted image can also be decoded and the display can be performed without deterioration. .

[0040]

As described above, according to the first aspect of the present invention, when a decoding error is detected in the error detecting procedure, it is possible to decode the frame by substituting the data of the adjacent decoded field. As a result, it is possible to quickly recover frames due to errors, minimize frame loss of images due to errors, and perform decoding continuously without affecting subsequent decoding. With the features of.

According to the second aspect, in the case where one frame is composed of the first field and the second field, when the field in which the error is detected is the second field, the first field decoded before that is the first field. It is possible to restore the frame of the field in which the error has occurred by substituting the data of (1), and it is possible to eliminate the influence of the next decoding by eliminating the discontinuity of the frame.

According to the third aspect, when the field in which the error is detected is the second field and the field decoded after the error is the second field, the error is detected in the second field decoded. Since the frame is restored by substituting the first field that was written, even if an error occurs and the next frame cannot be decoded, the frame can be formed immediately after the decoding, and the frame is lost. , And the effect on the next decoding can be minimized.

According to claim 4, when the field in which the error is detected is the second field which forms a frame with the first field after the error is detected in the first field, the decoding is performed. Secondized
By substituting the field for the first field in which an error is detected, the frame is restored, so that the frame can be formed continuously and the influence on the next decoding can be eliminated.

According to the fifth aspect, the field in which the error is detected is the first field, the field decoded after the error is detected is the second field, and the same as the first field in which the error is detected. If the frame is not formed, the decoded second field is replaced with the first field forming the same frame as the decoded second field to restore the frame, so that an error occurs, and Even if the field of is not decoded, the frame is restored immediately after being decoded, the loss of the frame can be minimized, and the influence on the subsequent decoding can be minimized. Has features.

According to the sixth aspect, since the lost frame is restored by substituting the frame restored in the fifth aspect for the frame of the field in which the error is detected, the loss of the frame is minimized. It has features such as being able to minimize the influence on the next decoding.

According to the seventh aspect, when a decoding error is detected by the error detecting means, the data of the adjacent decoded field is substituted for the field in which the error is detected to restore the frame. Features such as rapid frame restoration operation for errors, minimizing the loss of image frames due to errors, and eliminating the impact on subsequent decoding as frames are restored Have.

[Brief description of drawings]

FIG. 1 is a principle diagram of the present invention.

FIG. 2 is a block diagram of an embodiment of the present invention.

FIG. 3 is an operation flowchart of error recovery processing according to an embodiment of the present invention.

FIG. 4 is an operation explanatory diagram of error recovery processing according to an embodiment of the present invention.

FIG. 5 is an operation explanatory diagram of error recovery processing according to an embodiment of the present invention.

FIG. 6 is a diagram for explaining the structure of an interlaced image.

[Explanation of symbols]

1 Error Detection Procedure 2 Data Substitution Procedure 11 Decoding Device 12 Variable Length Decoding Section 13 Inverse Quantization Section 14 Frame Memory 15 Adder 16 Display Section 17 Error Recovery Control Section S1 _{-1 to} S1 _-12 Processing Steps

Claims (7)

[Claims] 1. A decoding method of an interlaced coded image, which decodes coded data generated by a coding method of coding an interlaced image in units of fields constituting a frame, wherein: An error detection procedure for detecting an error in the data, and when an error is detected in the error detection procedure, the data in the field adjacent to the error is detected, and the decoded field data of the field in which the error is detected is detected. A method for decoding an interlaced coded image, comprising a data substitution procedure for substituting for data. 2. The frame comprises a first field and a second field.
When the field in which an error is detected in the error detection procedure is the second field, the data substitution procedure substitutes the data of the first field decoded before that for the second field. The method for decoding an interlaced coded image according to claim 1, further comprising a first substitute procedure. 3. The frame comprises a first field and a second field.
If the field in which the error is detected in the error detection procedure is the second field and the field decoded after the error is the second field is an error, The decoding method of an interlaced coded image according to claim 1 or 2, further comprising a second substitution procedure for substituting the data of the second field, which is decoded after the detection, with the data of the first field. 4. The frame comprises a first field and a second field.
In the data substitution procedure, the field in which an error is detected in the error detection procedure is the first field, and the field decoded after the error is detected is the first field in which the error is detected. In the case of a second field forming a frame with a frame, a third substitution procedure for substituting the data of the decoded second field for the data of the first field is included. Or a method for decoding an interlaced coded image as described above. 5. The frame includes a first field and a second field.
In the data substitution procedure, the field in which the error is detected in the error detection procedure is the first field, the field decoded after the error is the second field, and the error is A fourth substitute for substituting the data of the decoded second field for the data of the first field forming the same frame as the decoded second field when the same frame as the detected first field is not formed. The method of decoding an interlaced coded image according to any one of claims 1 to 4, further comprising a procedure. 6. The data substituting procedure has a fifth substituting procedure for substituting the data of the frame formed by substituting by the fourth substituting procedure for the data of the frame including the first field in which the error is detected. The method of decoding an interlaced coded image according to claim 5, wherein. 7. A decoding device for interlaced coded images, which decodes coded data generated by a coding method for coding interlaced images in units of fields constituting a frame, Error detecting means for detecting an error in the data, when an error is detected in the error detecting procedure, the data of the field adjacent to the field in which the error is detected, and the decoded field of the field in which the error is detected is detected. An apparatus for decoding an interlaced coded image, comprising: data substitute means for substituting for data.