一种引入遗传算法的最大类间距离法在脑部图像分割中的运用  

A Method of Introducing Genetic Algorithm to Brain Image Segmentation Following OSTU

作　　者：李志勇[1] 张丽辉[2] 孙建洪[3] 

机构地区：[1]云南红河学院信息技术中心,云南蒙自661100 [2]云南师范大学教务处,昆明650092 [3]云南红河学院工学院计科系,云南蒙自661100

出　　处：《红河学院学报》2014年第2期30-32,共3页Journal of Honghe University

基　　金：红河学院博硕项目(XSS08010)

摘　　要：最大类间距离法已经被证明是一种较好的阈值化方法[1],但在用于双阈值或多阈值分割时存在实现效率较低等问题,而遗传算法高效的随机搜索能力正好能弥补其不足.根据脑部图像分为灰质、白质等三类的特点,通过把遗传算法引入最大类间距离法来实行对脑部图像的分割,实验验证取得了较好的分割效果.OTSU method[1]has been proved to be a better threshold method that use to image segmentation[2], but when the segmentation threshold is dual- or multi-, the segmentation efficiency become low. Genetic algorithm is an efficient random search algorithm, so introduce genetic algorithm to OSTU can improve the segmentation efficiency. According to the brain images are divided into three categories of material： gray, white, and the other. Through introduce genetic algorithm to OSTU to implement brain Image dual-threshold Segmentation. Experimental verification has achieved good segmentation results.

关 键 词：遗传算法 最大类间距离法 双阈值 脑部图像 分割 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]