A novel fast motion estimation method based on particle swarm optimization

Motion estimation is an important and computationally intensive task in video application. Block matching based fast algorithm reduce the computational complexity of motion estimation at the expense of accuracy. Fast motion estimation algorithms often assume monotonic error surface in order to speed up the computations involved in motion estimation. But search may trap into local minima resulting in motion estimates. In this paper, we propose a new fast motion estimation algorithm based on an improved particle swarm optimization (PSO). The method can overcome the weakness of being liable to local minima resulting through particle swarm sharing optimized information during searching. Thus, the estimation speed is prompted. By applying spatial correlation to particle initialization, the performance of search is also improved. Some experiments presented demonstrate the efficiency of proposed approach. The regularity and high parallelism of PSO make it feasible for VLSI implementation of video encoders.