Image analysis – Image compression or coding
Reexamination Certificate
1999-01-14
2002-02-05
Chen, Wenpeng (Department: 2624)
Image analysis
Image compression or coding
C348S584000, C370S487000, C725S032000
Reexamination Certificate
active
06345122
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates generally to a compressed picture data editing apparatus and method and recording medium for storing instructions to perform such editing and to store such edited data, and more particularly to a compressed picture data editing apparatus and method in which two compressed picture data streams are spliced to each other at an arbitrary splice point to generate one continuous compressed picture data stream, and a recording medium on which the continuous compressed picture data stream is stored.
Recently, digital transmission systems for moving pictures such as a digital broadcasting system, a visual telephone system, etc. have been developed. In such digital transmission systems in general, moving picture data is compressed according to the MPEG-2/Video method. MPEG is an abbreviation for the “Moving Picture coding Experts Group” of the ISO/IEC JTC1/SC29 (International Organization for Standardization/International Electrotechnical Commission, Joint Technical Committee 1/Sub Committee 29). An ISO 11172 standard is provided for an MPEG Phase 1, and an ISO 13818 standard is provided for an MPEG Phase 2. In these international standards, ISO 11172-1 and 13818-1 are provided for system multiplexing, ISO 11172-2 and 13818-2 are provided for video coding, and ISO 11172-3 and 13818-3 are provided for audio coding.
According to the MPEG-2/Video standard (ISO 13818-2), when decoding a compressed picture data input, in order to prevent an input buffer located at the front part of a decode circuit from underflowing or overflowing, the capacity of the input buffer should be considered, and the amount of compression during encoding should be controlled so that upon decoding the input buffer can accommodate the supplied data amount without underflowing or overflowing. When encoding according to the MPEG-2/Video standard, a virtual buffer for monitoring and controlling the amount of code that would be generated during decoding is defined as a VBV (Video Buffering Verifier) buffer, and the capacity of this VBV buffer is defined as VBV buffer size (vbv_buffer_size). The function of the VBV buffer will be further described with reference to
FIG. 14
, which depicts the amount of data stored in an input buffer provided upstream of a decode circuit when the buffer is supplied with input MPEG streams. The horizontal axis of
FIG. 14
represents time along which decoding times (t
101
, t
102
, t
103
, . . . ) for various pictures are shown while the vertical axis represents the amount of data (bit occupancy) stored in the input buffer. The input buffer will store the MPEG streams compressed according to the MPEG-2/Video Method sequentially at their respective bit rates.
During a decoding procedure, the data corresponding to a first picture is extracted from the VBV buffer by the decoder at a time t
101
. This time t
101
includes a delay (vbv_delay) from a time t
100
, the time at which the supply of the picture data is started, to allow for the data amount stored in the input buffer to reach an initial predetermined amount so that all of the data associated with the first few pictures can be loaded into the VBV buffer. Data representing the length of the delay (vbv_delay) is included as part of the MPEG stream in a picture layer. The time is represented by a clock of 90 kHz, for example.
During decoding, the amount of data corresponding to a picture to be extracted from the VBV buffer by the decode circuit to be decoded is the sum of the amount of data comprising the picture (picture_size), the amount of data comprising a picture start code (picture_start_code), and the amount of data comprising a sequence header (sequence_header). The amount of data resulting from the addition of picture_size, picture_start_code and sequence_header will be referred to as “image size (image_size)”.
After time t
101
and the extraction of the data of the first image to be decoded from the VBV buffer, the input buffer will be sequentially supplied with a succession of MPEG data streams at a predetermined bit rate corresponding to subsequent transmitted pictures. Also decoding of sequential pictures takes place at times t
102
, t
103
, . . . , t
n
, t
n+1
, . . . , &Dgr;DTS being a time interval between decode procedures. At each time t
102
, t
103
, . . . , t
n
, t
n+1
, . . . , an amount of data corresponding to the image size of each picture is extracted by the decode circuit from the VBV buffer for decoding. Thus, the input buffer provided upstream of the decode circuit will store the MPEG streams received sequentially at their respective bit rates while an amount of data corresponding to image size of each picture is extracted by the decode circuit at each of the decoding times (t
101
, t
102
, t
103
, . . . ).
If the total amount of data of the supplied MPEG streams exceeds the total amount of data corresponding to the image sizes of the pictures extracted from the VBV buffer at the decoding times, that is more data is added to the VBV buffer than is removed from it, the input buffer will eventually overflow when this excess data exceeds the size of the VBV buffer. If the total amount of data of the image sizes of the pictures extracted for decoding exceeds the total amount of data of the supplied MPEG streams, that is less data is added to the VBV buffer than is removed from it, the input buffer will underflow because more data will be taken out of the VBV buffer than will be input to it. To avoid overflowing or underflowing the VBV buffer, the MPEG-2/Video (ISO 13818-2) standard prescribes that when encoding MPEG streams, the capacity of an input buffer provided upstream of a decode circuit during decoding should be modeled as a VBV buffer size (vbv_buffer_size) to allow for control of the amount of encoded data produced during encoding. The model input buffer is thus designed to accommodate an appropriate amount of data during encoding so that during decoding, the amount of data contained in the VBV buffer will remain within a predetermined range.
Generally, at a broadcasting station or the like, in order to transmit a moving picture sequence that has been compressed according to the MPEG-2 system, often two or more sets of moving picture data are edited and spliced together to form one set of moving picture data. For example, at the broadcasting station, moving picture data of a movie may be edited to insert a commercial into the movie as shown in FIG.
15
. Each of the commercials includes an amount of data that allows the commercial to run for a short period of time. When conventionally performing such an edit, for real-time transmission and for convenience in editing, base-band moving picture data are not handled during editing. Rather, compressed MPEG streams are processed during editing without decoding.
First to third conventional alternatives for performing such editing will now be described. In each of the conventional alternatives, a first picture data stream compressed according to the MPEG-2/Video Method (“main stream” hereinafter) has spliced thereto a second picture data stream also compressed according to the MPEG-2/Video Method (“sub-stream” hereinafter) to generate one spliced data stream. The conventional first to third alternatives will be described with reference to
FIGS. 16
to
18
, which show the amount of data in the VBV buffer attributable to the main stream, the amount of data in the VBV buffer attributable to the sub-stream, and the amount of data in the VBV buffer attributable to the spliced stream, respectively.
First, referring to
FIG. 16
, the conventional first alternative will be described. As is shown in
FIG. 16A
, a splice-out point at which the main stream is to be spliced out is defined at as a time (t
111
). At this splice-out point, the supply of data to the VBV buffer from the main stream will end. In the first conventional alternative, a splice-in point of the sub-stream is defined at a time (t
121
). At this splice-in point the supply of the sub-stream to the VBV buffer will start, as shown in FIG.
Negishi Shinji
Tahara Katsumi
Yamato Atsushi
Yasuda Mikita
Chen Wenpeng
Frommer William S.
Frommer & Lawrence & Haug LLP
Kessler Gordon
Sony Corporation
LandOfFree
Compressed picture data editing apparatus and method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Compressed picture data editing apparatus and method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Compressed picture data editing apparatus and method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2969441