Apparatus and method for efficient conversion of DV (digital video) format encoded video data into MPEG format encoded video data by utilizing motion flag information contained in the DV data
First Claim
1. A video signal conversion apparatus for converting DV (Digital Video) encoded video data to MPEG (Motion Picture Expert Group) encoded video data, said DV data consisting of digital video data which have been encoded in a DV format and including motion flag data which specify for each of respective video data blocks whether interlaced-field mode DCT (Discrete Cosine Transform) processing or progressive-field DCT processing has been applied in encoding that block, the apparatus comprising:
- video decoding means (11) for decoding said DV encoded video data and for extracting said motion flag data from said DV encoded video data to obtain decoded video data formed of a stream of interlaced-field frames, and video encoding means (15) coupled to receive said decoded video data and said motion flag data, for executing MPEG format encoding of said decoded video data, wherein said video encoding means (15) includes processing mode selection means responsive to said motion flag data for adaptively selecting a mode of said MPEG format encoding, based upon said motion flag data.
1 Assignment
0 Petitions
Accused Products
Abstract
A video signal conversion apparatus converts DV-format encoded compressed video data to MPEG format data, by utilizing motion flag information contained in the DV data which specifies for each of respective DCT blocks whether interlaced-frame mode DCT processing or progressive-field mode DCT processing has been applied to the block when executing DV encoding. The motion flag information can be used by a procesing mode selection section to determine various MPEG encoding modes to be applied in units of MPEG macroblocks, such as selecting respective macroblocks to be subjected to interlaced-frame mode DCT processing by an orthogonal transform processing section when the set of motion flags corresponding to the blocks constituting that macroblock are judged to indicate no substantial amount of motion for the macroblock, and to select progressive-field mode DCT processing when a substantial amount of motion is judged to exist. The amount of MPEG processing can thereby be readily reduced, without deterioration of image quality.
84 Citations
22 Claims
-
1. A video signal conversion apparatus for converting DV (Digital Video) encoded video data to MPEG (Motion Picture Expert Group) encoded video data, said DV data consisting of digital video data which have been encoded in a DV format and including motion flag data which specify for each of respective video data blocks whether interlaced-field mode DCT (Discrete Cosine Transform) processing or progressive-field DCT processing has been applied in encoding that block, the apparatus comprising:
-
video decoding means (11) for decoding said DV encoded video data and for extracting said motion flag data from said DV encoded video data to obtain decoded video data formed of a stream of interlaced-field frames, and video encoding means (15) coupled to receive said decoded video data and said motion flag data, for executing MPEG format encoding of said decoded video data, wherein said video encoding means (15) includes processing mode selection means responsive to said motion flag data for adaptively selecting a mode of said MPEG format encoding, based upon said motion flag data. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13)
macroblock extraction means 21 coupled to receive said decoded video data, for extracting successive macroblocks from each of said frames, interlaced-field mode DCT block extraction means (23) for receiving and directly outputting each macroblock which is extracted by said macroblock extraction means 21, progressive-field mode DCT block extraction means (24) for receiving each macroblock which is extracted by said macroblock extraction means 21, separating each of said macroblocks into two half-macroblocks which are contained in a first field and in a second field of each of said frames respectively, and successively outputting said half-macroblocks, and conversion mode control means (22) controlled by said processing mode selection means for selecting each of said macroblocks of said decoded video data to be transferred directly by said interlaced-field mode DCT block extraction means (23) to said DCT processing means (25) for application of interlaced-field mode DCT processing when it is judged from said motion flag data that said macroblock exhibits a relatively small amount of motion, or to be transferred as successive halfblocks from said progressive-field mode DCT block extraction means (24) to said DCT processing means (25) for application of progressive-field mode DCT processing, when it is judged from said motion flag data that said macroblock exhibits a relatively large amount of motion.
-
-
4. The video signal conversion apparatus according to claim 1, wherein said video encoding means (15) includes motion prediction processing means (18) and said processing mode selection means comprises means for selectively applying progressive-field mode motion prediction processing by said motion prediction means to a macroblock extracted from a frame of said decoded video data when it is judged from said motion flag data that said macroblock exhibits a relatively large amount of motion and for applying interlaced-field mode motion prediction processing to said macroblock when it is judged from said motion flag data that said macroblock exhibits a relatively small amount of motion.
-
5. The video signal conversion apparatus according to claim 4, wherein said video encoding means (15) includes video memory means (148) for storing reconstructed video data frames which have been reconstructed from encoded video data, motion search means (43) and prediction error derivation means (44), and wherein said processing mode selection means comprises;
-
interlaced field mode referenced macroblock extraction means (140) for obtaining from said video memory means a reconstructed frame for use as a reference frame, and outputting successive reference macroblocks from said reference frame, progressive-field mode reference macroblock extraction means (141) for obtaining from said video memory means a reconstructed frame for use as a reference frame, and for separating said frame into first and second fields as first and second reference pictures, and sequentially outputting successive reference macroblocks extracted from said first field, then successive reference macroblocks extracted from said second field, interlaced-field mode object macroblock extraction means (143) for directly extracting successive macroblocks from a frame of said decoded video data, and for outputting said directly extracted successive macroblocks, progressive-field mode object macroblock extraction means (144) for extracting successive macroblocks from a frame of said decoded video data, and for separating each of said macroblocks into a first half-macroblock which is contained in a first field of said frame and a second half-macroblock which is contained in a second field of said frame, and for successively outputting said first and second half-macroblocks, and processing mode control means (16, 42) for selecting reference macroblocks produced from said interlaced-field mode reference macroblock extraction means (140) and a directly extracted macroblock which is produced from said interlaced-field mode object macroblock extraction means (143), to be supplied to said motion search means (43) and said prediction error derivation means (44) for applying interlaced-field mode motion prediction processing to said macroblock, when it is judged from said motion flag data that said macroblock exhibits a relatively small amount of motion, and for selecting reference macroblocks produced from said progressive-field mode reference macroblock extraction means (141) and a successive pair of halfblocks of an extracted macroblock which are produced from said progressive field mode object macroblock extraction means (144), to be supplied to said motion search means (43) and said prediction error derivation means (44) for applying motion prediction processing to said macroblock, when it is judged from said motion flag 15 data that said macroblock exhibits a relatively large amount of motion.
-
-
6. The video signal conversion apparatus according to claim 1, wherein said mode selection means selects, for each of successive frames of said decoded video data, interlaced-field mode DCT processing and motion prediction processing when it is judged from said motion flag data that the image contents of a frame exhibit a large amount of motion and progressive-field mode DCT processing and motion prediction processing when it is motion judged from said motion flag data that the image contents of said frame exhibit a small amount of motion.
-
7. The video signal conversion apparatus according to claim 6, wherein said mode selection means comprises picture formation means (51) coupled to receive said decoded video data and controlled by said processing mode selection means (52) for outputting each interlaced-field frame of said decoded video data unchanged, to be subjected to interlaced-field mode DCT processing and motion prediction processing, when it is judged from said motion flag data that said interlaced-field frame exhibits a small amount of motion and for outputting said each interlaced-field frame as two consecutive fields, to be subjected to progressive-field mode DCT processing and motion prediction processing, when it is judged from said motion flag data that said interlaced-field frame exhibits a relatively large amount of motion.
-
8. The video signal conversion apparatus according to claim 1, wherein said video encoding means (15) includes motion prediction processing means (18) comprising motion search means (343) for comparing a macroblock extracted from a frame of said decoded video data with successive macroblocks of a reference frame within a specific search range in said reference frame, and wherein s aid processing mode selection means comprises search range control means (61) for operating on said motion search means to set said search range as a narrow range when it is judged from said motion flag data that said macroblock exhibits a relatively small amount of motion and to set said search range as a wide range when it is judged from said motion flag data that said macroblock exhibits a relatively large amount of motion.
-
9. The video signal conversion apparatus according to claim 1, wherein said video encoding means (15) includes means for applying intra-coding and means for applying inter-coding to respective video data macroblocks, and wherein said processing mode selection means comprises means (81) for transferring a macroblock of an interlaced-field frame of said decoded video data to be subjected to said intra-coding when it is judged from said motion flag data that said macroblock exhibits a relatively large amount of motion, and for transferring said macroblock to be subjected to said inter-coding when it is judged from said motion flag data that said macroblock exhibits a relatively small amount of motion.
-
10. The video signal conversion apparatus according to claim 1, wherein said processing mode selection means judges whether a macroblock exhibits a large or a small degree of motion based upon the states of respective motion flags of luminance signal blocks of said macroblock.
-
11. The video signal conversion apparatus according to claim 1, wherein said processing mode selection means judges that a macroblock exhibits a large amount of motion when at least a predetermined number of respective motion flags of luminance signal blocks of said macroblock each indicate a condition of large amount of motion.
-
12. The video signal conversion apparatus according to claim 1, wherein said processing mode selection means judges that a macroblock exhibits a small amount of motion when at least one of respective motion flags of two color difference signal blocks in said macroblock indicates a condition of small amount of motion.
-
13. The video signal conversion apparatus according to claim 1, wherein said processing mode selection means judges that a macroblock exhibits a large amount of motion when at least a predetermined number of respective motion flags of four luminance signal blocks of said macroblock and also at least one of respective motion flags of two color difference signal blocks of said macroblock each indicates a condition of large amount of motion.
-
14. A method of converting DV (Digital Video) encoded video data to MPEG (Motion Picture Expert Group) encoded video data, said DV data consisting of digital video data which have been encoded in a DV format and including motion flag data which specify for each of respective video data blocks a DCT processing mode which has been applied in DV encoding of a block, whether interlaced-field mode DCT (Discrete Cosine Transform) processing or progressive-field DCT processing has been applied in encoding that block, the method comprising:
-
decoding said DV encoded video data to obtain decoded video data formed of a stream of interlaced field frames, and extracting said motion flag data from said DV encoded video data, and executing MPEG format encoding of said decoded video data, by adaptively selecting a mode of said MPEG format encoding, based upon a judgement of said motion flag data. - View Dependent Claims (15, 16, 17, 18, 19, 20, 21, 22)
-
Specification