Three-layer Fast Image Matching Algorithm Research Based on Evolutionary Algorithm

Abstract

A three-layer fast image matching algorithm based on Evolutionary Algorithm is proposed. It has some new features: 1A strategy from coarse matching to fine matching is adopted. Large numbers of non-matching points will be firstly eliminated by performing coarse matching with circular and cross templates, then a whole template is applied to confirming the final position to reduce the calculation workload; 2) two mutation strategies are proposed: low probability mutation strategy for the early mutation; and high probability strategy for the late mutation to enhance the diversity of population. The experimental results demonstrate that the performance in this paper outperforms that of other evolutionary algorithms in terms of the quality of the final solution, its stability is better and its computational cost is lower than the cost required by the correlation method and the circular method.