A novel search window selection scheme for the motion estimation of HEVC systems

This paper presents a search window selection approach that can reduce the computational complexity of motion estimation in HEVC systems. In this scheme, the size of the search window is decided only based on the Motion Vector Predictor (MVP) of the Largest Coding Unit and is inherently adaptive to the characteristic of the video content. The experimental results show that the average accessed pixel in the search window is 81.4 % reduced compared to the conventional full-search method and 45.3 % decreased compared to the prior art. In addition, compared to the HEVC Test Model (HM), the proposed approach only results in 0.067 dB drop in average peak signal-to-noise ratio (PSNR).