基于遗传算法的目标检测定位方法  

TWO TARGET DETECTION AND LOCALIZATION METHODS BASED ON GENETIC ALGORITHMS

作　　者：侯格贤[1] 吴成柯[1] 

机构地区：[1]西安电子科技大学通信工程学院

出　　处：《航空学报》1997年第6期681-686,共6页Acta Aeronautica et Astronautica Sinica

基　　金：国防预研基金

摘　　要：将目标的检测定位归结为一类组合优化问题，利用遗传算法的进化策略，采用改进的遗传算法，提出了２种基于遗传算法的目标检测定位方法：基于遗传算法的相关匹配定位法和图象序列目标检测定位法。并给出了在可见光图象序列上的实验结果。The method of target detection and localization is a key technique for automatic target recognition and tracking system. The paper is directed towards the problem with reducing target detection and localization to a sort of combinatorial optimization problem. By using the evolution strategy of genetic algorithms and improving techniques on genetic algorithms, two methods of target detection and localization are presented. The experiments conducted on visible band image sequences are reported.

关 键 词：目标检测定位 相关匹配 遗传算法 目标识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]