Motion estimation algorithm and architecture survey
Shengyang Chen · Applied and Computational Engineering · 2023
Motion estimation is a key part of video temporal characteristics analysis including video filter and compression. Increasing requirements for high-quality video image processing have confirmed the necessity of researching motion estimation. Current motion estimation methods, including high-efficient fast motion estimation algorithms and hardware architecture designs, achieved different performances from different perspectives. In this paper, we focus on analyzing typical fast algorithms and hardware architecture designs for motion estimation, comparing the used methods, and pointing out the basic ideas of motion estimation research schemes. In terms of theoretical implications, the research contributed to a more comprehensive reference for fast motion estimation algorithms and their hardware architecture designs.