基于竞选算法的特征点图像匹配研究  

Research on Feature Points Matching Based on Election Campaign Optimization Algorithm

作　　者：梅文宝 吕文阁[1] 陈凯[1] 

机构地区：[1]广东工业大学机电工程学院,广东广州510006

出　　处：《机电工程技术》2016年第4期7-10,共4页Mechanical & Electrical Engineering Technology

基　　金：广东省科技计划项目(编号:02311690227499035)

摘　　要：针对在特征点图像匹配过程中,存在搜索时间长,匹配速度慢的问题,将特征点图像匹配过程转化成求最优解问题,把竞选算法应用到特征点图像匹配的搜索匹配过程,寻找最优解,从而优化匹配过程,提高匹配的性能,以PCB板定位实验可证,在同等匹配精度的前提下,把竞选算法应用到特征点图像匹配过程,可以大大减少匹配时间,加快匹配速度。The algorithms of image matching based on feature points mostly have the disadvantages of long time of search and slow matching velocity. In order to optimize matching process and improve the matching performance,the problem of image based on feature points will be transformed into a optimization problem,search feature points through Election Campaign Optimization Algorithm,then find the global optimal solution. The image matching of PCB location experiment shows that the feature points image matching based on Election Campaign Optimization Algorithm greatly increase the matching velocity without lossing matching accuracy.

关 键 词：特征点 图像匹配 竞选算法 最优解 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]