Genetic block matching algorithm for video coding

Abstract

Genetic algorithms (GAs) recently have been successfully applied to perform block-based motion estimation. It is shown that the performance of the GA-based motion estimation algorithms nearly approaches that of the full search algorithm (FSA). However, the computational complexity of the existing GA-based algorithms is too high to be used in practice. In this paper, a lightweight genetic search algorithm (LGSA) is proposed. It can be seen from the simulation results that the performance of the proposed LGSA is not only as good as that of the FSA, but the computational complexity is also much lower than that of the FSA and the other existing genetic motion estimation algorithms.