Block Matching Algorithm Based on Particle Swarm Optimization for Motion Estimation

Abstract

In this paper, based on particle swarm optimization (PSO), we propose a fast block matching algorithm for motion estimation (ME) and compare the algorithm with other popular fast block-matching algorithms for ME. A real-world example shows that the block matching algorithm based on PSO for ME is more feasible than others. Moreover, the initial values of parameters in PSO are empirically discussed, since they directly affect the computational complexity. Thus, an improved PSO algorithm for ME is empirically given to reduce computational complexity.