Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1998-10-01
2002-07-16
Lee, Richard (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
Reexamination Certificate
active
06421385
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an apparatus for video signal conversion, and to a corresponding video signal conversion method, for converting a compressed digital video signal into a compressed digital video signal in a different compression format. In particular, the invention relates to a method and apparatus whereby motion flags which are contained in the compressed digital video signal prior to conversion are effectively utilized for converting the video signal.
2. Description of the Prior Art
In recent years, with the increasing popularity of multimedia, considerable research has been executed in various directions for video on-demand systems, whereby video images can be viewed from a television receiver or a personal computer whenever desired.
A group which includes the assignees of the present invention has developed and put into practical application a system, for use as a video on-demand system, whereby images which are acquired by a digital video camera are edited and then converted to a compressed format digital video signal, to be distributed to personal computers etc.
In general, a digital video camera records a compressed digital video signal in the DV format, which is a standard that has been established for digital video equipment. The DV format was established as a standard in 1996, for application to video cassette recorders, based on the “Specifications of Consumer-Use Digital VCRs” (HD Digital VCR Conference, 1996), whereby image compression is achieved by a combination of DCT (Discrete Cosine Transform) processing to reduce spatial image redundancy within each frame of a digital video signal and variable-length encoding to reduce code redundancy.
With video data in accordance with the DV format, as shown in diagram (a) of
FIG. 12
, one macroblock of a video signal frame consists of four luminance signal blocks which are arrayed along the horizontal direction as an elongated rectangular array, with each luminance signal block consisting of an array of 8×8 pixel values, and two color difference signal blocks (CR, CB) which, in a finally displayed picture each correspond in size and position to the set of four luminance signal blocks, with this arrangement of pixel values within a frame being referred to as the 4:1:1 color component format. Also, as shown in diagram (b) of
FIG. 12
, 27 of these macroblocks constitute a super block, with a 5×10 set of the super blocks constituting one complete frame of the digital video signal.
A DV format video camera outputs video information in units of interlaced fields, with a {fraction (1/60)} second period between successive fields. When the amount of image motion within a frame is small, then each frame is formed by combining two successive fields, so that the frame period is {fraction (1/30)} second, and DCT processing is applied to each of such frames as an interlaced combination of two fields, with such processing being referred to as interlaced-field mode DCT processing. On the other hand, when the image motion within a frame is large, then the frame is processed as two successive fields, i.e., DCT processing is separately applied to each of the two fields of that frame, with such processing being referred to in the following as as progressive-field mode DCT processing. Since it is possible that image motion may occur within only a limited region, the selection of frame DCT mode or field DCT mode is executed adaptively in units of blocks of a frame. When DCT processing is applied to each of the four luminance signal blocks and two color difference signal blocks of a macroblock, respective motion flags corresponding to these six blocks are inserted into the code which is generated by compressing the digital video signal, with these motion vectors respectively indicating for each block whether field DCT mode or frame DCT mode has been applied to that block. A motion flag takes the logic value “1” if the amount of motion detected for the corresponding block is large, so that progressive-field mode DCT processing has been assigned to the block, and takes the value “0” if the amount of motion detected for the corresponding block is small, so that interlaced-frame mode DCT processing has been assigned to the block. These motion flags are subsequently referred to when decoding the compressed DV format digital video signal.
In order to distribute digital video data that has been compressed in accordance with the DV standard, to personal computers, etc., it is necessary to convert the video data to the MPEG-1 or MPEG-2 compressed code format. This conversion is generally executed by decoding the DV standard video data to recover a non-compressed video signal consisting of successive frames, and then applying compression processing in accordance with the MPEG standard to the non-compressed digital video signal.
The MPEG-1 or MPEG-2 compression standards are widely applied to video signals which are to be processed by personal computers. Each of these is a standard whereby spatial image redundancy within each frame is reduced by applying DCT transform processing, then applying variable-length encoding to reduce code redundancy. In addition, inter-frame redundancy is reduced by applying motion compensation. For that reason, the amount of code which is generated by MPEG compression encoding is reduced to ⅙ of the amount of that is generated by DV compression encoding, so that the MPEG code can easily be transmitted via a network.
MPEG-1 is described in detail in IOS/IEC 11172-2 “Information technology—Coding of moving pictures and associated audio for digital storage media at up to about 1.5 Mbit/s—Part 2: Video”, while MPEG-2 is described in IOS/IEC 13818-2 “Information technology—Generic coding of moving pictures and associated audio information—Part 2: Video”
With MPEG format video data, as shown in diagram (c) of
FIG. 13
, one macroblock is formed of four luminance signal blocks (each formed of 8×8 pixel values) arranged in a square array, and two color difference signal blocks (CR, CB) which correspond in position to the set of four luminance signal blocks, with this being referred to as the 4:2:0 color component format. In terms of respective amounts of data, since each color difference signal block consists of 8×8 pixel values it is equivalent to one luminance signal block, however in terms of a finally displayed picture (after interpolation of color difference values), each color difference signal block of a macroblock corresponds in size and position to the set of four luminance signal blocks of that macroblock. A set of macroblocks arrayed along the horizontal scanning direction of a frame constitutes one slice, as shown in diagram (b), with a plurality of slices constituting one picture, as shown in diagram (a).
With MPEG encoding, intra-coding (i.e. direct conversion to sets of DCT coefficients) or inter-coding (i.e., motion prediction encoding, and conversion of resultant prediction error values to DCT coefficients) of the digital video signal may be applied.
With the MPEG-1 method, encoding in units of fields is not executed. This enables the processing speed to be high, however since a period of {fraction (1/60)} second occurs between the times at which the respective images of two successive fields are captured by a video camera, when these two fields are combined into a single field-interlaced frame and such a frame is directly encoded, deterioration of the resultant reproduced image quality will occur whenever there is rapid motion within the image expressed by a frame.
This problem is reduced with the MPEG-2 encoding method. In that case, progressive-field mode encoding or interlaced-field mode encoding can be adaptively selected. With progressive-field mode encoding applied to an entire frame, the two fields constituting the frame are separated, and the entire contents of each field are separately encoded (by intra-coding or inter-coding). Alternatively, only certain macroblocks of the frame can be adaptively selected
Komiya Daisaku
Uenoyama Tsutomu
Lee Richard
Loew Hauptman Gilman & Berner, LLP
Matsushita Electric - Industrial Co., Ltd.
LandOfFree
Apparatus and method for efficient conversion of DV (digital... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for efficient conversion of DV (digital..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for efficient conversion of DV (digital... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2825608