基于遗传算法的二维最大类间方差图像分割算法  

Algorithm of for 2D Maximum Between-Cluster Image Segmentation Based on GA

作　　者：王军强[1] 高雅[1] 杨数强[1] 

机构地区：[1]洛阳师范学院物理与电子信息学院,河南471000

出　　处：《微计算机应用》2011年第8期24-30,共7页Microcomputer Applications

基　　金：洛阳师范学院青年科研基金项目(10000872)

摘　　要：为解决传统二维最大类间方差(0tsu)阈值分割算法处理图像时计算复杂度高、实时性差、易受噪声干扰等问题,本文将遗传算法应用到二维Otsu法中,提出一种基于遗传算法的最大类间方差法的灰度图像分割算法。二维Otsu算法考虑了图像的灰度信息及邻域空间的相关信息,以保证图像分割的精度;利用遗传算法则能提高运算速度。因此,基于遗传算法的最大类间方差法的灰度图像分割算法兼有二者优点,不仅提高运算速度而且能保证图像分割精度。The traditional two- dimensional maximum between -cluster variance method （0tsu） threshold segmentation algorithm has the high computation complexity, poor real - time and noise sensitivity in processing image. In this paper, to solve these problems, a 2 -D Otsu method based on an improved genetic algorithm（GA） was improved, and lay the emphases on the gray image segmentation method that combine the genetic algorithm and the Otsu method. The 2 - D Otsu method not only consider the gray information of the image, but also the related information of neighbor space to ensure the accuracy of image segmentation. The result shows that ： the proposed algorithm have the two advantage, more calculating speed and more precision.

关 键 词：图像分割 二维OTSU法 遗传算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]