Binary Block Motion Estimation Using an Adaptive Search Range Adjustment Technique

Abstract

This study proposes an adaptive search range adjustment algorithm to enhance the effectiveness of current binary block motion estimation methods. This is because the binary block ME approaches, which perform FS by using different criterion from the generally used sum of absolute differences (SAD), has a significant drawback in PSNR. The combination of the proposed algorithm with the binary block motion estimation is thought to enhance 1) motion estimation accuracy and 2) time efficiency. The findings showed that the proposed combination, compared to the conventional binary block motion estimation method, significantly increased object visual qualities and time efficiency.