红河学院学报
紅河學院學報
홍하학원학보
JOURNAL OF HONGHE UNIVERSITY
2014年
2期
30-32
,共3页
遗传算法%最大类间距离法%双阈值%脑部图像%分割
遺傳算法%最大類間距離法%雙閾值%腦部圖像%分割
유전산법%최대류간거리법%쌍역치%뇌부도상%분할
Genetic Algorithm(GA)%OTSU%dual-threshold%brain image%segmentation
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。
最大類間距離法已經被證明是一種較好的閾值化方法[1],但在用于雙閾值或多閾值分割時存在實現效率較低等問題,而遺傳算法高效的隨機搜索能力正好能瀰補其不足。根據腦部圖像分為灰質、白質等三類的特點,通過把遺傳算法引入最大類間距離法來實行對腦部圖像的分割,實驗驗證取得瞭較好的分割效果。
최대류간거리법이경피증명시일충교호적역치화방법[1],단재용우쌍역치혹다역치분할시존재실현효솔교저등문제,이유전산법고효적수궤수색능력정호능미보기불족。근거뇌부도상분위회질、백질등삼류적특점,통과파유전산법인입최대류간거리법래실행대뇌부도상적분할,실험험증취득료교호적분할효과。
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.