Reducing Motion Estimation Time with Skin Detection
Sajad Shirali-Shahreza, Sajad Shirali-Shahreza · 2009
Motion estimation (ME) is one of the key parts of video compression algorithms. But, motion estimation and computation of motion vectors (MVs) are very time consuming. In this paper, we propose a method for reducing the cost of motion estimation process. During this process, a series of candidate blocks should be searched to find the best motion vectors. In our method, we compare the skin parts of two blocks before comparing all pixel pairs of the two blocks. Having a preprocessing phase, the skin part comparison is performed quickly. This method provides a parameter that can be used to create a balance between the processing time and the motion estimation accuracy.