改进的FCM与局部信息相结合的图像分割  

IMAGE SEGMENTATION BASED ON COMBINATION OF IMPROVED FCM AND LOCAL INFORMATION

作　　者：黄宁宁[1] 贾振红[1] 余银峰[1] 杨杰[2] 庞韶宁 

机构地区：[1]新疆大学信息科学与工程学院,新疆乌鲁木齐830046 [2]上海交通大学图像处理与模式识别研究所,上海200240 [3]新西兰奥克兰理工大学知识工程与开发研究所

出　　处：《计算机应用与软件》2011年第8期97-99,129,共4页Computer Applications and Software

基　　金：科技部国际科技合作项目(2009DFA12870)

摘　　要：传统的EM算法和FCM算法分割精度低,时间消耗大。为解决以上不足,提出了一种基于EM、FCM和KCN三种算法相结合的全新的图像分割算法。该算法有较好的分割精度。考虑到图像会受到噪声的干扰,在改进算法的基础上又引入图像的局部信息。首先利用图像的局部信息重塑图像的灰度直方图,增强了像素的类间散布性和类内紧凑性,然后让改进的算法在重塑图像上执行。实验结果表明,该算法具有很好的分割效果和较强的抗噪性能。The traditional EM algorithm and FCM algorithm have many problems,such as great amount of calculation and slow operation speed.In this paper we propose a novel image segmentation algorithm based on the combination of three algorithms of EM,FCM and KCN for resolving the above problems.The algorithm has higher segmentation accuracy.Considering the noise the image might be affected,based on the improvement of the algorithm we also introduce local information of the image.In the first step,the gray-value histogram of given image is reshaped using local information of the image to enhance inter-class interspersion and intra-class compactness of the pixel.In the second step,the improved algorithm is performed on the reshaped image.Experimental results show that the proposed method has good segmentation effect and good performance in noise resistance.

关 键 词：期望最大 KOHONEN网络 模糊C均值聚类 局部信息 图像分割 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]