基于微粒群优化和粗糙熵的图像分割算法  

Image Segmentation Algorithm Based on Particle Swarm Optimization and Rough Entropy

作　　者：王明[1,2] 黄海峰 何峰[1,2] 左文艳 

机构地区：[1]江苏大学计算机科学与通信工程学院,江苏镇江212013 [2]镇江市高等专科学校电子与信息工程系,江苏镇江212003 [3]镇江高等职业技术学校电气工程与自动化系,江苏镇江212016

出　　处：《计算机工程》2011年第14期228-230,共3页Computer Engineering

基　　金：江苏省产业信息化重点基金资助项目(1633000004)

摘　　要：提出一种基于微粒群优化(PSO)的边界区域粗糙熵的阈值图像分割算法。该算法采用边界粗糙熵作为图像分割的评价标准,利用优化领域的PSO功能把图像分割问题转化为优化问题。实验结果表明,该方法使用PSO算法避免了早期大量熵的计算,相对于分块大小的敏感性较小,得到较好的分割效果,并且能提高计算速度,是一种实用有效的图像分割方法。The image threshold segmentation algorithm based on the Particle Swarm Optimization（PSO） combined with the rough entropy based on boundary region is presented.The algorithm adopts the rough entropy based on boundary region as the valuation standard of image segmentation and converses image segmentation problem into an optimization problem and fully utilizes PSO function in the optimization field.Experimental results show that the proposed method can not only obtain the perfect performance of segmentation but also greatly improve the speed of computation,it avoids a great deal of entropy calculation for the use of PSO and the sensibility of the algorithm to the partition-size image sub-piece is low,it is a practical and effective method of image segmentation.

关 键 词：图像分割 微粒群优化 边界地区 粗糙熵 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]