免疫遗传算法最大类间方差图像分割法研究  

Research on Immune Genetic Algorithms for Maximum Interclass Variance Image Segmentation

作　　者：王学忠[1] 李美莲[1] WANG Xue-zhong;LI Mei-lian(College of Electronic and Electrical Engineering, Anhui Sanlian University, Hefei 230601,China)

机构地区：[1]安徽三联学院电子电气工程学院

出　　处：《佳木斯大学学报（自然科学版）》2019年第5期736-738,817,共4页Journal of Jiamusi University：Natural Science Edition

基　　金：安徽三联学院科研基金项目(PTZD2019026);安徽省高校自然科学研究重点项目(KJ2018A0601)

摘　　要：图像分割是图像处理技术的一个重要的环节,传统采用GA-Otsu算法分割图像容易陷入局部最优、获得的最佳阈值不够稳定,图像分割不够清晰,伴有分割不足或过当等问题。针对传统GA-Otsu方法分割图像存在不足的问题,提出一种基于免疫的GA-Otsu算法(IGA-Otsu算法),基本思路是在传统GA-Otsu算法上引入免疫算子和浓度自适应调节,有效地增加了种群结构的复杂性,避免早熟。利用该方法可使获得的图像最佳阈值比较稳定,有效地避免了早熟,使分割的图像也更加清晰。Image segmentation is an important part of image processing technology. Traditional GA-Otsu algorithm is easy to fall into local optimum, the best threshold obtained is not stable, and the image segmentation effect is not good, which will be accompanied by inadequate or inappropriate image segmentation. An improved GA-Otsu algorithm (IGA-Otsu) is proposed to overcome the shortage of image segmentation in GA-Otsu method. The basic idea is to introduce immune operator and concentration adaptive adjustment into traditional GA-Otsu algorithm, which effectively increases the complexity of population structure and avoids premature maturity. Using this method, the best threshold of the image can be more stable, the prematurity can be avoided effectively, and the segmented image can be clearer.

关 键 词：免疫遗传算法 阈值 图像分割 最大类间方差 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]