基于种内协同克隆选择的Otsu图像分割算法  

Image threshold segmentation method based on CSACS

作　　者：童成意 

机构地区：[1]长沙师范学院电子信息工程系,长沙410100

出　　处：《计算机工程与应用》2013年第24期161-164,共4页Computer Engineering and Applications

基　　金：湖南省自然科学基金重点项目(No.13JJA002)

摘　　要：传统二维Otsu算法存在计算复杂度高、实时性差等缺点。针对这一不足,受生物群体成员间协作关系的启示,对克隆免疫算法进行改进,提出了一种基于种内协同的克隆选择算法(Clonal Selection Algorithm based on Cooperation within Species,CSACS),将其与克隆选择算法(Clonal Selection Algorithm,CSA)进行对比测试,将其应用于二维Otsu图像分割。测试实验表明:该算法能加快收敛速度,具有较好的实时性,且分割效果较为理想。Traditional twodimensional Otsu＇ s method has the disadvantages of high computational complexity, poor realtime performance. In order to improve its efficiency, inspired by the collaborative relationships among group members, a Clonal Selection Algorithm based on Cooperation within Species（CSACS） is proposed. CSACS is compared with Clonal Selection Algorithm （CSA）, and is used to image segmentation. The experimental results show that：the algorithm can accelerate the con vergence speed, has good realtime ability, and its segmentation effect is very well.

关 键 词：克隆选择算法 协同进化 图像分割 OTSU 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]