An Adaptive Motion Estimation Algorithm Based on Object Segmentation

In order to reduce the computational complexity of the video compression algorithm and improve the accuracy of rebuild images, a new adaptive motion estimation coding algorithm is proposed. First use a new adaptive genetic algorithm to segment the video object, identify the adaptive genetic mechanisms, the early generation of individuals, selection operator and judgments of the end, and draw the boundaries. Next adopt the overall removal within the region, use a new adaptive motion estimation algorithm based on spatial-temporal correlation to outside the region. The result can reduce the search scope and the number of search points. Compared with diamond search algorithm, experimental results showed that the time of searching is reduced to 4.2 ms and PSNR is increased about 1.6 dB by the new algorithm.