Multi-Frame Motion Estimation for Video Compression

The current frame is divided into multiple blocks, and for each block, only selected reference frames are examined rather than all five reference frames. This segmentation of the search space allows the system to maintain coding quality while reducing the overall processing time by avoiding redundant examinations of all reference frames for every block.

Instead of performing exhaustive motion estimation on all five reference frames for every block, the patent applies partial action by examining only a subset of reference frames that are most likely to contain the best match. This partial examination approach maintains sufficient coding quality while significantly reducing processing time.

The reference frame search space is segmented by examining only specific reference frames for each block rather than all five frames uniformly. This selective segmentation reduces computational power requirements while maintaining coding result quality by focusing computational resources on the most promising reference frames.

Different blocks in the current frame may use different numbers of reference frames based on local characteristics. This local quality approach allows the system to allocate computational power efficiently, using more reference frames for blocks that benefit from it while using fewer for blocks where it is unnecessary, thus reducing overall computation while maintaining coding results.