Survey on block based pattern search techniques for motion estimation

Motion Estimation (ME) and Motion Compensation (MC) techniques are used for video compression which eliminates the temporal redundancy of video sequence. For motion estimation, Block Matching Algorithms (BMA) has been widely used among the various motion estimation algorithms because of its simplicity and efficiency. In BMA, search pattern with different shape or size has very important impact on search speed and distortion performance. Among various BMA, fixed pattern search algorithms are the most famous; these algorithms reduce the computational expenses significantly less, while PSNR close to Full Search (FS) algorithm. In this paper we have reviewed in brief already implemented various block based pattern search techniques and present their respective advantages and disadvantages and also review their performances based on the number of searching points and PSNR.