Image analysis – Image compression or coding – Including details of decompression
Reexamination Certificate
1997-10-31
2001-12-11
Johnson, Timothy M. (Department: 2623)
Image analysis
Image compression or coding
Including details of decompression
C348S425200, C375S240270
Reexamination Certificate
active
06330365
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a decoding method and a decoding apparatus that are employed when Information is transmitted through a transmission line and is stored in a storage device.
BACKGROUND OF THE INVENTION
In recent years, with the advance of an information-oriented society, transmission of moving pictures without the barriers of time and distance has been increasingly demanded. The time has come when digital techniques are made practicable in earnest, and it is possible to record and reproduce moving pictures in a recording device and to transmit the moving pictures over a long distance via a communication network In addition, not only in a communication field but in a broadcasting field, transmission using digital techniques and adoption of a coding system have been realized.
Generally, moving pictures and audio signals of digital signals are large in amount of coding. Therefore, in order to record and transmit these moving pictures and audio signals with good efficiency, it Is necessary to utilize high-efficiency coding techniques, and various coding apparatus and decoding apparatus have been already manufactured by way of trial.
As examples of application, there are a video CD (compact disk) in which digital moving pictures are recorded in a CD, and a DVD which records digital moving pictures at higher image quality and for a longer time than those of the video CD.
For a decoding apparatus that reproduces these video CD and DVD, trick play, such as fast forward play and fast reverse play, is indispensable. In order to realize the trick play, there is a method corresponding to one that is described in the international standard, called MPEG (moving picture experts group), “Information Technology—Coding of moving pictures and associated audio for digital storage media at up to about 1.5 Mbits/s” (ISO/IEC11172-2). A description will be given of a conventional method for decoding digital moving pictures using the above-described method.
First, a method for coding digital moving pictures and a bitstream in the MPEG are described.
In the MPEG, assuming that digital moving pictures comprise a series of video frames
700
as shown in FIG.
5
(
a
), a video frame group
500
, which is called a sequence, is coded. This sequence is usually divided into a series of video frame groups
600
at about 0.5 second each, which are called GOP (group of pictures), to be coded
The GOP comprises I picture, P pictures and B pictures, for example, as schematically shown in FIG.
5
(
b
). The I picture is an Intra-Picture that is coded with only data in the video frame, the P picture is a Predictive Picture that is predicted and coded from data of the I picture or P picture of two frames before, and the B picture is a Bidirectionally Predictive-Picture that is predicted and coded from data of the I and P pictures or the P pictures before and after the B picture.
FIG.
6
(
a
) shows a structure of each picture. The picture comprises a continuous region or more having a belt shape on a picture plane, which is called a slice. The slice comprises a block
800
or more of 16 lines long and 16 pixels broad, which is called a macroblock.
As shown in FIGS.
6
(
b
) and
6
(
c
), the macroblock comprises a plurality of blocks of 8 lines long and 8 pixels broad, which are called blocks. For example, the macroblock comprises blocks of four luminance signals and respective chrominance signals of two systems (Cb, Cr).
As described above, the bitstream has a hierarchical structure. The sequence, the GOP, the picture and the slice, i.e., bitstreams constituting higher hierarchies of the hierarchical structure, include start codes for uniquely identifying these bitstreams on the bitstream, respectively. Further, the bitstreams have regions that retain hiearchical information, called headers and extensions, having coded information of the bitstreams constituting the respective hierarchies. For example, as shown in
FIG. 7
, these data are arranged to constitute a bitstream
1000
.
In the hierarchy of the macroblocks and the lower hierarchy, there is information, such as a macroblock address increment that shows how many macroblocks each macroblock is away from the macroblock which was coded last time by, a macroblock type that represents predictive mode information of the macroblock, which was selected at coding, a quantizer scale that represents a quantization step, a motion vector that is used for motion compensation, a coded block pattern that shows which blocks are coded and are present in the bitstream, and coded DCT (discrete cosine transform) coefficient information.
At this time, variable length codes are used in coding the information of the hierarchy constituted by the macroblocks and the lower hierarchy. By allocating a shorter code to the information appearing more frequently, the information of the hierarchy constituted by the macroblocks and the lower hierarchy, which occupies the large portion of the bitstream, is coded efficiently.
In reality, as shown in
FIG. 8
, the bitstream is divided into packets
40
of appropriate lengths. Similarly, audio signals that are coded separately from the moving pictures are divided into packets. In addition, a packet header
41
that retains identification information is added to the head of each packet
40
, for identifying the information of the packet These packets are multiplexed to form the bitstream
1000
. The information in each packet
40
that has been originally included in the bitstream
1000
is called a payload region
42
.
A description is given of a fast forward play method and a fast reverse play method of the bitstream.
In a case of normal play, the bitstream is all reproduced and all the pictures are decoded to be displayed. In a case of fast forward play, however, all the bitstream is transmitted to a decoding apparatus, and the I pictures are selected from the bitstream and decoded in the decoding apparatus to be displayed. Alternatively, only the I pictures are selectively transmitted to the decoding apparatus, and these pictures are decoded in the decoding apparatus to be displayed.
In the decoding apparatus, using a code table that shows relationships between codes and symbols, such as picture data, corresponding to the codes, decoding is performed while collating bitstreams that are successively input with the code table. When the decoding apparatus finds a symbol that is not defined in the code table, decoding of a picture with the symbol is stopped immediately, and the picture is subjected to error concealment processing as error processing.
As an example of error concealment processing, concerning a picture that has not been decoded by stopping, a content of the previous picture that has been decoded is copied to a portion of the picture that is not accomplished, thereby accomplishing a video frame, and the video frame is displayed in place of the picture that has not been decoded by stopping. Then, processing of skipping to a header of a next picture is performed.
As another example of error concealment processing, a picture that has not been decoded by stopping is not displayed at all, and processing of skipping to a header of a next picture is only performed.
There are a case in which one of these error concealment processings is always performed, and a case in which the processing is adaptably switched. Similarly, when the presence of the header and extension does not satisfy a specified syntactic rule, an error is considered to occur, and the same processing as described above is performed.
However, when practically using the above-described method for fast forward play, the capability to parse the bitstream is insufficient, and the selection of the I pictures is complicated. Therefore, without transmitting all bitstream fragments including all the bitstream
1000
to the decoding apparatus as shown in FIG.
9
(
a
), only bitstream fragments
300
,
310
,
320
, . . . comprising the packets
40
, each fragment being considered to include a specified bitstream, such as the I picture, are selective
Watabe Akihiro
Yasuda Makoto
Johnson Timothy M.
Matsushita Electric - Industrial Co., Ltd.
Parkhurst & Wendell L.L.P.
LandOfFree
Decoding method and apparatus using bitstreams and a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Decoding method and apparatus using bitstreams and a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Decoding method and apparatus using bitstreams and a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2584240