基于基因进化分支树算法的图像分割研究  

Infrared image segmentation based on gene evolutionary branch tree algorithm

作　　者：王宝红[1] 李宏升[1] 吕臻 季钢 

机构地区：[1]黄淮学院,河南驻马店463000 [2]河南省通信管理局,河南郑州450008 [3]驻马店供电公司,河南驻马店463000

出　　处：《激光与红外》2013年第8期939-942,共4页Laser & Infrared

摘　　要：针对图像分割的特点,采用基因进化分支树算法。首先构造基因,对基因的节点进行数学运算生成基因分支;接着在基因建树中,采用基于特征的构建基因树法,通过节点顺序来构建树,对分支采取最小支持项方法进行剪支和增支;最后把基因分支周围领域内的像素合并,图像中的所有的特征信息都分布在不同的区域中,给出了算法流程。实验仿真结果显示本文算法对图像分割效果连续,性能指标好。Aiming at the characteristics of infrared image segmentation, the gene tree algorithm is proposed. Firstly the gene is constructed, gene branch is generated through mathematical operations for gene node; Then in the genetic contribution, the gene tree is constructed based on the characteristics, a tree branch is built by the nodes order, the branches are sheared and increased by the minimum support method. Finally the pixels near the gene branches are in- corporated, all the characteristic information of the images distributes in different areas, and the algorithm flow is given. Simulation results show this algorithm segments continuous infrared image, and has good performance.

关 键 词：基因树 结构进化 分割 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]