Hybrid Particle Swarm Optimization And Its Application In Image Matching

With the increasingly widespread application of image matching technology, more and more new requirements appear, and how to improve the matching efficiency is a hot research problem, to meet the requirements of practical applications, on the one hand the calculation of similarity measure can be simplified, on the other hand the optimization algorithm can be found. Proposed in recent years, PSO is simple and has good global search ability, which has began to be used in many fields, but the research for image matching is not more. And PSO has its own drawbacks, easily falling into local optimal, the improvement of PSO is also a hot research problem. So taht, to simplify the calculation of similarity measure and improve PSO and carry out the research of image matching based on PSO has important significance and research value.Based on the research of traditional image matching algorithms, the maximum cross-correlation algorithm is simplified. The basic principle of PSO is studied deeply, and the causes of leading to premature convergence of PSO are analyzed, and the standard PSO is improved, and a hybrid particle swarm optimization HPSO is proposed. In order to use the good global search capability of HPSO, an image matching algorithm based on HPSO is designed. Combining the characteristics of image matching, the group is designed and the fitness function is choosed, and the detailed design of the algorithm is given. The parameter settings of the experiment are determined through lots of experiments and empirical values, and the algorithm is simulated and analyzed.Simulation results show that HPSO is better than the standard PSO in global search capability, and using HPSO for image matching is feasible, which can improve the matching efficiency and is an effective image matching algorithm. The study in this paper for improving the efficiency of matching and the theoretical study and application of PSO has a certain value.The experimental parameters are obtained by experiment, without effective theoretical basis, the method of parameter settings and the improvement and application expansion of PSO will be researched in the future.