| 摘要 |
A method provides a fast motion search in advanced video signal coding systems based on a reference-frame prediction and a block-mode prediction so that a motion search of each block mode and each reference frame is not required. A reference frame prediction fp, spaced from the current frame (22) by "p" number of frames, is determined by: p = min(n-1, po + max(a,b,c,d)); wherein po is a pre-chosen positive integer, n is the total number of reference frames (12, 14, 16, 18, 20), wherein A, B, C, and D are image blocks adjacent to a searched block, and wherein the reference image blocks have been chosen from reference frames fa, fb, fc and fd. The search is conducted within frames fo to fp, which is a subset of all the n reference frames (12, 14, 16, 18, 20), so that the total computational burden is significantly decreased with respect to prior art searches. The mode frequency prediction is based on the frequencies of the block modes: F<0> = alpha . min(FmA, FmB, FmC, FmD); wherein alpha is a positive parameter less than 1.0. The block-mode selection is then conducted using the mode-frequency prediction. Each mode m among all the M possible modes will be considered if Fm is greater than or equal to F<0>. If Fm is less than F DEG then mode m will be skipped during the motion search. <IMAGE> |
