WO2006008673A1

WO2006008673A1 - A method and apparatus for decoding encoded data

Info

Publication number: WO2006008673A1
Application number: PCT/IB2005/052204
Authority: WO
Inventors: Xiaotian Guo
Original assignee: Koninklijke Philips Electronics N.V.
Priority date: 2004-07-13
Filing date: 2005-07-04
Publication date: 2006-01-26
Also published as: CN1722830A

Abstract

The invention provides an error processing method and apparatus for decoding data. The data comprises a group of pictures which includes multi-frame correlated pictures. Firstly, detecting if the received the group of pictures is continuous; secondly, determining the type of the lost picture if the group of pictures is not continuous; lastly, processing the group of pictures based on the determined type of the lost picture and the corresponding error correction strategy. The types of pictures include pictures that have to be referred to when decoding other pictures and pictures that do not have to be referred to when decoding other pictures.

Description

A METHOD AND APPARATUS FOR DECODING ENCODED DATA

FIELD OF THE INVENTION The present invention relates to a method and apparatus for decoding encoded data, particularly to a method and apparatus for decoding encoded picture data.

BACKGROUND OF THE INVENTION

MPEG (Moving Picture Expert Group) standard provides a technical solution of compressing/encoding digital video signal. This solution employs predictive encoding technique to correspondingly correlate each frame of picture in the original data to be encoded, and to determine the corresponding picture type of each frame of picture based on the correlation properties thereof.

There are generally three types of video picture frames according to the MPEG standard, i.e., I-frame, P-frame and B-frame. An I-frame is an intraframe encoded picture, and is encoded by using only the information of its own. A P-frame is an interframe forward-direction predictive encoded picture, and is encoded by using motion compensation prediction according to the previously encoded I-frame or P-frame. A B-frame is an interframe bi-direction predictive encoded picture, and is encoded by using motion compensation prediction according to the previously encoded I-frame or P-frame and the subsequent I-frame or P-frame.

Accordingly, during decoding, an I-frame picture is decoded by using only its own data, and a P-frame picture is decoded by using data in the I-frame or P-frame that has been decoded previously. Similarly, a B-frame picture is decoded by using data in the previous and subsequent I-frame and/or P-frame.

During encoding video pictures of the MPEG standard, all the picture frames could be divided into several groups of pictures (GOP). A typical group of pictures includes 15 frames of picture, wherein there is at least one I-frame picture. Generally speaking, in the sequence of encoding and decoding, the I-frame is the first frame of picture in the group of pictures. Meanwhile, a picture header is usually added before the content of each frame of picture.

Fig. Ia is a schematic diagram of the content structure of the picture header in accordance with the MPEG standard. The picture header includes three important codes, the first one

110 is the picture_start_code, for recognizing the start of a frame of picture, that is, notifying the decoder of the coming of a frame of new picture; the second code 120 is the temporal_reference, for indicating the position of the frame picture in the group of pictures thereof, and eventually for indicating the sequence of the frame of picture when displaying; the third code 130 is the picture_coding_type, for indicating the type of a frame of picture, for example, I-frame, P-frame or B-frame. The other codes in the picture header are not closely relevant to the present invention, therefore they are not shown in the figure.

The digital video program satisfying the MPEG standard could be stored in a storage medium, such as a DVD (Digital Versatile Disc). The program could be decoded by a decoder complying with the MPEG standard during playback so as to obtain the original pictures before encoding for displaying. However, due to the quality problems of the

DVDs, such as scratches, the headers of some picture frames, the picture_start_codes therein in particular, will be lost, and thus causing the reduction of the playing quality, such as mosaic, or even playing failure.

The digital video program satisfying MPEG standard could also be transmitted, such as broadcast, via a special channel. In the receiving side, the program could be decoded by a decoder complying with the MPEG standard so as to obtain the original pictures before encoding for displaying. Likewise, due to the interference of the channels, the receiving side will fail to receive the headers of some pictures , and if the picure_start_codes are lost, the playing quality will be reduced, for example, mosaic.

Fig. 1 is a schematic diagram illustrating the damage of the loss of the header. In the figure, the picure_start_codes of No. 3 frame, No. 4 frame and No. 5 frame are damaged, in this case, since the pictures of No. 3 to No. 5 frames cannot be detected, the decoder will ignore Nos. 3 to 5 frames and decode No. 6 frame immediately. The decoding will be re-established on the basis of No. 0 frame, i.e., the I- frame. Without the occurrence of the above-mentioned damage, the decoding re-establishment of No. 6 frame should be based on No. 3 frame. Thus, it can be seen that owing to the loss of No. 3 frame, the decoding of No. 6 frame is performed on a wrong basis and the decoding quality will be inevitably reduced. Further, since No. 6 frame is also the basis of the decoding re-establishment of the subsequent Nos. 7-14 frames, such decoding error will be spread to the subsequent frames and thereby expanding the range of the influence of the error; accordingly, the decoding quality will be inevitably greatly reduced and mosaic will be clearly observed.

The U.S. Patent 5838265 (Inventor: Dirk Adolph, Date of Patent: Nov. 17, 1998) has provided a technical solution to resolve the above-mentioned problem. The Patent mentioned the above-mentioned problem and it copied some important parameters in the header of the picture into the header of the slice when encoding, so that if the header of the picture is lost, some important parameters of the picture could be obtained from the header of the slice for decoding.

The drawback of the Patent is that although it mentioned the problem of loss of the header of the picture, especially the loss of the picture_start_code, it failed to propose a technical solution for how to determine the loss of a frame of picture. The technical solution adopted in the Patent was to copy some important parameters in the header of the picture, but this will increase the data redundancy after encoding. Besides, the current encoded data satisfying the MPEG standard has to be re-encoded after adopting the technical solution, and this will bring about a large amount of work.

Therefore, a new method and apparatus for decoding encoded data is needed, which could recognize the lost picture frames during decoding and perform the corresponding processing, thereby the decoding could be carried out smoothly and the decoding quality could be increased.

OBJECT AND SUMMARY OF THE INVENTION

One of the objects of the present invention is to eliminate the drawback of the current video decoder and to provide a new method and apparatus for decoding encoded video data, which could recognize the lost picture frames during decoding and perform the corresponding processing, thereby the decoding could be carried out smoothly and the decoding quality could be increased.

The present invention provides an error processing method for decoding data, with the data comprising a group of pictures, the group of pictures includes multi-frame correlated pictures. The method comprising: firstly, detecting if the received group of pictures is continuous; secondly, if the group of pictures is not continuous, determining the type of the lost picture; finally, processing the group of pictures according to the determined type of the lost picture and the corresponding error correction strategy. Wherein the types of pictures include pictures that have to be referred to when decoding other pictures and pictures that do not have to be referred to when decoding other pictures.

In the method provided by the present invention, the continuity of the pictures in a group of pictures could be detected, and further, the type of the lost picture could be determined.

Therefore, the error correction strategy adopted will be more effective, and therefore, even if the picture header is lost, the decoding could be performed well.

In an embodiment of the present invention, the pictures in the group of pictures are correlated with each other in time sequence. By comparing the time sequence parameter of the received picture with the expected possible time sequence parameters, it could be determined if there is continuity between the received picture and the previously received frame of picture. The expected possible time sequence parameters are obtained by counting the different types of pictures in the received the group of picture respectively.

In another embodiment of the present invention, if the lost picture is one that has to be referred to when decoding other pictures, the lost picture is recovered on the basis of the picture of the same type received prior to the lost picture. Thus, the picture could still be referred to when decoding the subsequent pictures and the wrong picture will not be referred to, accordingly, the expansion of error is avoided.

The present invention further provides an error processing apparatus for decoding data. The data comprises a group of pictures, the group of pictures includes multi-frame correlated pictures. The apparatus comprises: detecting means for detecting if the received group of pictures is continuous; determining means for determining the type of the lost picture if the group of pictures is not continuous; and processing means for processing the group of pictures according to the determined type of the lost picture and its corresponding error correction strategy.

The present invention also provides a decoder for decoding data. The data comprises a group of pictures, the group of pictures includes multi-frame pictures. The decoder comprises: header decoding means for decoding the header information of the pictures; content decoding means for decoding the content information of the pictures; and error processing means for detecting errors according to the header information of the pictures and performing the corresponding processing. The error processing means comprises: detecting means for detecting if the received group of pictures is continuous; determining means for determining the type of the lost picture if the group of pictures is not continuous; and processing means for processing the group of pictures according to the determined type of the lost picture and the corresponding error correction strategy.

The other objects and achievements of the present invention will become obvious on the basis of the following descriptions and the claims with reference to the figures, and meanwhile, a more comprehensive understanding could be acquired on the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention will be explained in detail by way of examples with reference to the figures, wherein:

Fig. Ia is a schematic diagram of the content structure of a picture header satisfying the

MPEG standard;

Fig. 1 is a schematic diagram showing the damage of the loss of the header; Fig. 2 is a system block diagram of an error processing apparatus according to an embodiment of the present invention;

Fig. 3 is a system block diagram of a decoder according to an embodiment of the present invention;

Fig. 4 is a schematic diagram showing the operation of determining the continuity of pictures according to an embodiment of the present invention; and

Fig. 5 is a flow chart showing error processing during decoding according to an embodiment of the present invention.

In all the figures, the same reference numerals represent similar or same features and functions.

DETAILED DESCRIPTION OF THE INVENTION Fig. 2 is a system block diagram showing an error processing apparatus according to an embodiment of the present invention. The error processing apparatus 200 is used for decoding data, the data includes a group of pictures. The group of pictures comprises multi-frame correlated pictures. These pictures are correlated with each other during encoding and are divided into different types according to correlation properties thereof. The types of the pictures include pictures that have to be referred to when decoding other pictures and pictures that do not have to be referred to when decoding other pictures.

The apparatus 200 comprises detecting means 210, determining means 220 and processing means 230. The apparatus 200 could also selectively comprise creating means 240.

The apparatus 200 could be a part of a decoder (not shown in the figure), wherein a part of the functions of the apparatus could be realized by software. For those skilled in the art, the above-mentioned means included in the apparatus 200 could be realized by various current means, as long as their combination could achieve the functions of the present invention.

The detecting means 210 is used for detecting if the received group of pictures is continuous. The detecting means judges if a new group of pictures is received based on the picture header of the received group of pictures, and may also judge if the picture is continuous with the picture received prior to it according to the header of the received picture. The detecting means 210 further comprises counting means 212 and comparing means 218. The counting means 212 is used for counting the pictures of different types in the received group of pictures respectively to obtain the expected possible time sequence parameters. The operation principle of the counting means is as shown in the following Fig. 4. The comparing means 218 compares the time sequence parameter of the received picture with the expected possible time sequence parameters in the counting means 212 to determine if the picture is continuous with the picture received prior to it.

The determining means 220 is used for determining the type of the lost picture if the group of pictures is not continuous. If the detection result from the detecting means 210 is that the received picture is not continuous with the picture received prior to it, the determining means determines the type of the lost picture based on the picture type table created by the creating means 240 and the time sequence parameters of the two frames of pictures (i.e., the currently received picture and the picture received prior to it).

The processing means 230 is used for carrying out corresponding processing on the group of pictures according to the type of the lost picture and its corresponding error-correction strategy. The processing means adopts the corresponding error processing strategy to process the lost picture and/or its subsequent pictures according to the type of the lost picture determined by the determining means 220. The processing means could also selectively comprise ignoring means 232, discarding means 234 and recovering means 236. If the lost picture is one that does not have to be referred to when decoding other pictures, the ignoring means 232 ignores this lost picture, i.e., ignores the error and takes no further action. If this lost picture is one that has to be referred to when decoding other pictures, the ignoring means 232 could ignore the lost picture, and the discarding means 234 discards all the pictures that are subsequent to this lost picture in the group of pictures. Of course, another option, in this case, is that the recovering means 236 recovers the lost picture according to the picture of the same type received prior to the lost picture.

The creating means 240 is used for creating a picture type table according to the received group of continuous pictures, the picture type table indicates the type of each frame of picture in the group of pictures and the corresponding time sequence parameter. The types of the group of continuous pictures and the corresponding time sequence parameters could be obtained from the detecting means 210. The operation flow of the apparatus 200 is shown in the following Fig. 5, wherein the operation flow of detecting means 210 is shown in Fig. 4.

Fig. 3 is a system block diagram of a decoder according to an embodiment of the present invention. The decoder comprises header decoding means 310, error decoding 200 and content decoding means 330.

The header decoding means 310 is substantially the same as the prior art header decoding means, for decoding sequence header, GOP header, picture header and slice header, etc., to acquire the corresponding header information.

The structure and function of the error decoding apparatus 200 are as shown in the above Fig. 2. The apparatus detects the received picture according to the header information acquired by the header decoding means 310, and judges if it is continuous with the picture received prior to it, and if not, the type of the lost picture is further determined and the corresponding error correction measure is taken according to the type of the lost picture. If the lost picture is a picture that does not have to be referred to when decoding other pictures, the error could be ignored. If the lost picture is a picture that has to be referred to when decoding other pictures, the error could also be ignored, and all the pictures subsequent to the lost picture in the group of pictures will be discarded instead of being sent to the content decoding means 330 to be decoded. Of course, if the lost picture is a picture that has to be referred to when decoding other pictures, another option is to recover the lost picture based on the picture of the same type received prior to the lost picture, and the recovered picture could be used as reference for the content decoding means 330 to decode its subsequent pictures.

The content decoding means 330 is substantially the same as the prior art content decoding means, which comprises variable length coding (VLD) decoder, inverse quantization (IQ), inverse discrete cosine transform (IDCT), motion compensation, etc. The content decoding means 300 performs selective content decoding under the control of the error processing means. If the lost picture determined by the error processing apparatus 200 is the one that has to be referred to when decoding other pictures and the corresponding error correction measure is to recover the lost picture based on the picture of the same type received previously, the content decoding means 330 could decode the subsequent pictures according to the recovered picture.

For those skilled in the art, the above-mentioned means included in the means 300 could be realized by various current means, as long as their combination could achieve the functions of the present invention. The functions of some of the means could be realized by software.

The operation flow of the means 300 is shown in the following Fig. 5.

Fig. 4 is a schematic diagram showing the operation of determining the continuity of the pictures according to an embodiment of the present invention. The schematic diagram reflects the operation process of the detecting means 210.

In the present embodiment, the picture is from an encoded picture data stream, and the encoded data, which complies with the MPEG standard, comprises a plurality of groups of pictures with each including 15 frames of picture. These pictures are correlated with each other during encoding and are divided into different types, such as I-frame, P-frame and B-frame, according to their correlation properties. I-frame and P-frame are pictures that have to be referred to when decoding other pictures, while B-frame is one that does not have to be referred to when decoding other pictures.

The encoding sequence of the 15 frames in the group of pictures is I, B, B, P, B, B, P, B, B, P, B, B, P, B, B, and the corresponding temporal_ reference (TR) is 2, 0, 1, 5, 3, 4, 8, 6, 7, 11, 9, 10, 14, 12, 13. The TR value is encoded into the picture header, representing its real displaying sequence during displaying.

In the present embodiment, as shown in Fig. Ia, every time a picture_start_code is detected, it means that a frame of new picture is received, and the type of each frame of received picture could be obtained through the picture_coding_type in the header of the picture.

When the first frame of picture of a new group of pictures is received, i.e., an I-frame, an IP_counter is set to the corresponding TR value of this frame, and the TR is 2 here. Thereafter, every time a frame of picture of the group of pictures is received, the value of the IP_counter is added by 1 automatically regardless of the type of the picture. The value of the IP_counter is an expected possible time sequence parameter, which could be used for predicting the TR value of the received I-frame or P-frame.

When the first B-frame of the group of pictures is received, a B_counter is set to the corresponding TR value of the frame, and the TR is 0 here. Thereafter, every time a frame of picture of the group of pictures is received, the value of the B_counter is added by 1 automatically regardless of the type of the picture. The value of the B_counter is another expected possible time sequence parameter, which could be used for predicting the TR value of the received B-frame.

For a continuous group of pictures, by comparing the values of the IP__counter and the B_counter with the TR value of the received picture, it could be easily found that the TR value should be equal to the value of the IP_counter if the picture is an I-frame or P-frame, or the TR value should be equal to the value of the B_counter if the picture is a B-frame.

Thus, if the TR value equals neither the value of the IP_conter nor the value of the

B_counter, it indicates that the received picture is not continuous with the frame of picture received prior to it, that is to say, the group of pictures is discontinuous and there is a picture lost during decoding, wherein a very important reason is that the picture_start_code of the lost picture could not be detected.

In the MPEG standard, the first frame of picture in each group of pictures is generally an I-frame. If the first frame of picture in a group of pictures, i.e., the I-frame, is lost, the default TR value of the I-frame, e.g. 2, could be assigned to the IP_counter.

In addition, especially when beginning to decode a encoded data stream, if a continuous group of pictures is detected in the flow, a picture type table could be created based on the header of all the pictures in the group of pictures, and the picture type table indicates the type of each frame of picture in the group of pictures and its corresponding time sequence parameter. Please refer to Table 1 for the content of the table.

Generally speaking, since all the groups of pictures in an encoded data stream have the same picture type sequence, the creating process could be performed only once during decoding the encoded data, while the obtained picture type table could be stored temporarily for use in determining the type of the lost picture when decoding the subsequent groups of pictures. The table exists in the form of a linked list, but it could also exist in other forms as long as the same function is achieved.

The picture type table could also be added to the code stream, such as the user_data() of the MPEG standard, at one time according to the encoding principle during encoding, thus the table could be directly obtained from the code stream when decoding and does not need to be re-created. As for encoded data program complying with the DTV (Digital TV) standard, the table could be inserted into the program in the form of an application for carousel, for example, being broadcast once every 10 seconds, so that the user could receive the picture type table at any time during the program proceeding.

Fig. 5 is a schematic flow chart showing the error processing during decoding according to an embodiment of the present invention. The flow could be applied to the decoder as shown in Fig. 2.

Firstly, receiving a group of encoded picture data that complying with the MPEG standard (step S510), the encoded data includes a plurality of groups of pictures, and each group of pictures includes a plurality of frames (e.g. 15 frames) of pictures which are correlated with each other during encoding and are divided into different types, such as I-frame,

P-frame and B-frame, according to their correlation properties. Wherein I-frame and P-frame are pictures that have to be referred to when decoding other pictures, while B-frame is the picture that does not have to be referred to when decoding other pictures. A GOP header is added to each group of pictures during encoding, the header includes a group_start_code code, and by acquiring the group_start_code, the coming of a new group of pictures could be learnt.

Secondly, acquiring the header information of a picture (step S520). As shown in Fig. Ia, the header includes: a picture_start_code code for recognizing the coming of a frame of picture; a temporal_reference code for indicating the position of the frame of picture in the group of pictures and eventually for indicating the sequence of the frame of picture in display; and a picture_coding_type code for indicating the type of a frame of picture, such as I-frame, P-frame or B-frame.

Afterwards, detecting if the picture is continuous with the received previous frame of picture according to the content in the picture header information (step S530). One scheme of the detecting process is as shown in the above Fig. 4, that is, comparing the TR value in the picture header with its expected possible time sequence parameters, such as IP_counter and/or B_counter. If the TR value equals the IP_counter or B_counter, the picture is continuous with the frame of picture prior to it, and then performing content decoding on the received picture (step S 580).

If the detection result of step S530 indicates that the picture is not continuous with the received previous frame of picture, determining the type of the picture lost at discontinuity (step S540). The determining process is to search for the type of the lost picture in the picture type table as shown in Table 1 by using the TR value of the picture and the TR value of the picture received in the previous frame. For example, if the TR value of the currently received picture is 3, while the TR value of the picture received in the previous frame is 1, it could be determined from Table 1 that the lost picture is a P-frame with the TR value of 5.

Subsequently, adopting the corresponding error correction measure according to the type of the lost picture. If the lost picture is a B-frame (step S550), the error could be ignored, that is, the lost picture could be ignored (step S560) and content decoding is performed on the currently received picture (step S580) since a B-frame is a picture that does not have to be referred to when decoding other pictures. Of course, it is also possible to recover the lost B-frame by means of the current error correction measure.

If the lost picture is not a B-frame (step S550), i.e., it is an I-frame or P-frame, a certain error correction measure could be adopted to recover the lost picture (step S 570) since an

I-frame or P-frame is a picture that has to be referred to when decoding other pictures, for example, extrapolating on the basis of the picture information of the previous two P-frames to obtain the content of the lost frame, or interpolating on the basis of the content of the previous P-frame and the I-frame in the subsequent group of pictures to obtain the content of the lost frame. Then, performing content decoding on the currently received picture according to the content of the recovered picture (step S580). Another simple error correction measure is to ignore the lost picture and meanwhile discarding all the pictures that are subsequent to the lost picture in the group of pictures. In this case, the currently received picture as well as its subsequent pictures that belong to the group of pictures are all discarded without any content decoding, and upon the coming of the header of the next group of pictures, returning to step S510 to start a new flow.

Lastly, after step S580, judging if processing on the group of pictures is finished (step S590). The judgment is made on the basis of the type of the header of the picture after content decoding. If this header is still a header of a frame of picture, the processing on the group of pictures is not finished and it should return to step S520 to acquire the header information of the picture. If this header is a header of a group of pictures, the processing on the group of pictures is finished and a new group of pictures has come, but the decoding of the data is not finished (step S592) and it should return to step S510 to receive another group of encoded picture data that complies with the MPEG standard. If the header includes a sequence_end_code, the processing on the group of pictures is finished (step S590) and meanwhile the decoding of the data is finished (step S592), then the entire decoding flow ends.

Since in an encoded data stream the header of the group of pictures and the header of each of the pictures included in the group of pictures are arranged in sequence, the above-mentioned flow could be circularly performed until all the encoded data has been decoded.

In the above-mentioned flow, since the continuity of the pictures could be detected, the decoding of the subsequent pictures of the lost picture will not be based on a wrong reference picture even if the header, especially the picture_stard_code, of the picture is lost.

To be more specific, since the type of the lost picture could be determined, different error correction measures could be adopted for different types of the pictures, thus the advantage of the present invention is very obvious.

In addition, the encoded data satisfying the MPEG standard is taken as an example in the embodiment of the present invention, however, the method and apparatus of the present invention is applicable to any kind of encoded data, as long as the encoded data includes a plurality of groups of pictures and each group of pictures includes a plurality of frames of pictures, and meanwhile, the pictures are correlated with each other during encoding and are divided into different types according to their correlation properties.

Although the present invention is described above in conjunction with its specific embodiments, for those skilled in the art it is obvious to make many substitutions, modifications and variations. Therefore, when such substitutions, modifications and variations fall into the spirit and scope of the appended claims, they should be included in the present invention. It should be noticed that the preceding embodiments are only illustrative and not limiting the present invention, therefore those skilled in the art could design many alternative embodiments without departing from the scope of the appended claims. In the claims, the word "comprise" does not exclude the existence of the elements or steps that are not listed, and the word "a" or "an" used before the elements does not exclude the existence of a plurality of such elements. The present invention could be realized by hardware that comprises several particular elements or by a computer appropriately programmed.

Claims

What is claimed is:

1. An error processing method for decoding data, with said data comprising a group of pictures, said group of pictures comprising multi-frame correlated pictures, said method comprising the steps of: detecting if the received group of pictures is continuous; determining the type of the lost picture if said group of pictures is not continuous; and processing said group of pictures according to the determined type of the lost picture and the corresponding error correction strategy.

2. The method as claimed in claim 1, wherein said pictures are correlated with each other in time sequence, and said detecting step comprises comparing the time sequence parameters of the received pictures with the expected possible time sequence parameters.

3. The method as claimed in claim 2, wherein said expected possible time sequence parameters are obtained by respectively counting the different types of pictures in the received group of pictures.

4. The method as claimed in claim 1, wherein said types of pictures comprises pictures that have to be referred to when decoding other pictures and pictures that do not have to be referred to when decoding other pictures.

5. The method as claimed in claim 4, wherein if said lost picture is determined as a picture that does not have to be referred to when decoding other pictures, said processing step comprises ignoring said lost picture.

6. The method as claimed in claim 4, wherein if said lost picture is determined as a picture that has to be referred to when decoding other pictures, said processing step comprises: ignoring said lost picture; and discarding all the pictures subsequent to said lost picture in said group of pictures.

7. The method as claimed in claim 4, wherein if said lost picture is determined as a picture that has to be referred to when decoding other pictures, said processing step comprises recovering said lost picture according to the picture of the same type received prior to said lost picture.

8. The method as claimed in claim 1, further comprising the step of creating a picture type table based on said group of pictures if said group of pictures is continuous, wherein said picture type table is used to indicate the type of each frame of pictures in said group of pictures and its corresponding time sequence parameter.

9. The method as claimed in claim 1, wherein said data complies with the MPEG (Moving Picture Expert Group) standard.

10. The method as claimed in claim 1 or 9, wherein said types of pictures comprise I-frame, P-frame and B-frame.

11. An error processing apparatus for decoding data, with said data comprising a group of pictures, said group of pictures comprising multi-frame correlated pictures, said apparatus comprising: detecting means for detecting if the received group of pictures is continuous; determining means for determining the type of the lost picture if said group of pictures is not continuous; and processing means for processing said group of pictures according to the determined type of the lost picture and its corresponding error correction strategy.

12. The apparatus as claimed in claim 11, wherein said pictures are correlated with each other in time sequence, and said detecting means comprises comparing means for comparing the time sequence parameters of the received pictures with the expected possible time sequence parameters.

13. The apparatus as claimed in claim 12, wherein said detecting means further comprises counting means for respectively counting the different types of pictures in the received group of pictures to obtain the expected possible time sequence parameters.

14. The apparatus as claimed in claim 11, wherein said types of pictures comprise pictures that have to be referred to when decoding other pictures and pictures that do not have to be referred to when decoding other pictures.

15. The apparatus as claimed in claim 14, wherein if said lost picture is determined as a picture that does not have to be referred to when decoding other pictures, said processing means comprises ignoring means for ignoring said lost picture.

16. The apparatus as claimed in claim 14, wherein if said lost picture is determined as a picture that has to be referred to when decoding other pictures, said processing means comprises: ignoring means for ignoring said lost picture; and discarding means for discarding all the pictures subsequent to said lost picture in said group of pictures.

17. The apparatus as claimed in claim 14, wherein if said lost picture is determined as a picture that has to be referred to when decoding other pictures, said processing means comprises recovering means for recovering said lost picture according to the picture of the same type received prior to said lost picture.

18. The apparatus as claimed in claim 11, further comprising creating means for creating a picture type table based on said group of pictures if said group of pictures is continuous, wherein said picture type table is used to indicate the type of each frame of pictures in said group of pictures and its corresponding time sequence parameter.

19. A decoder for decoding data, with said data comprising a group of pictures, said group of pictures comprising multi-frame correlated pictures, said decoder comprising: header decoding means for decoding the header information of said pictures; content decoding means for decoding the content information of said pictures; and error processing means for detecting errors and performing corresponding process according to the header information of said pictures, said error processing means comprises: detecting means for detecting if the received group of pictures is continuous; determining means for determining the type of the lost picture if said group of pictures is not continuous; and processing means for processing said group of pictures according to the determined type of the lost picture and its corresponding error correction strategy.