混合粒子群优化法及其在图像匹配中的应用研究.pdfVIP

PSO PSO PSO PSO PSO PSO PSO PSO PSO (HPSO) HPSO HPSO HPSO HPSO HPSO PSO HPSO HPSO PSO Abstract 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