产生Tuned模板的基因扰动优化法研究与分析  

A Research and Analysis of Gene Disturb Optimization Method Which Produce Tuned Template

作　　者：郑肇葆[1] 郑宏[2] 

机构地区：[1]武汉大学遥感信息工程学院,武汉市珞喻路129号430079 [2]武汉大学电子信息工程学院,武汉市珞珈山430072

出　　处：《武汉大学学报（信息科学版）》2013年第10期1139-1141,1157,共4页Geomatics and Information Science of Wuhan University

基　　金：国家863计划资助项目(2009AA122002)

摘　　要：提出一个基于基因优化产生Tuned模板的新方法——基因扰动优化法(GDO法)。首先在初始模板群体中,根据求得的适应度值较好的少数几个模板,确定为优秀模板;然后对优秀模板中的每个基因分别进行扰动,根据扰动前后适应度值增加(或减少),确定每个优秀模板的两个优秀基因;最后,对优秀基因的数值引入扰动,得到一组新模板,进入下一代的计算,如此不断地优化得到最佳Tuned模板。经理论分析和对真实图像分类的实验证明,GDO法是一个很有潜力的优化方法。This paper presents a new method that is based on genetic optimization which produce tuned template, called genedisturb optimization method （referred GDO method）. Firstly, identified the ＂excellent template＂ according to a few templates of better fitness value which is obtained in the initial templates group. Then disturbing excellent template for each gene respectively, according to the value of fitness before and after the disturbance increases （or decreases） to determine two ＂ex- cellent genes＂ about each excellent template. Finally, disturb the numerical value of the excellent genes to get a new set of templates, then go into the next generation of calculation, constantly opti- mize the best tuned template. The theory analysis and the real image classification experiment proved that GDO method is a very potential optimization method.

关 键 词：基因扰动优化 Tuned模板 图像纹理分类 

分 类 号：P237.3[天文地球—摄影测量与遥感]