基于粗集与FCM的快速图像分割的改进方法  

Improved Method of Fast Image Segmentation Based on Rough Sets and FCM

作　　者：张朝全[1] 周绍景[2] 梁颖[3] 

机构地区：[1]江西理工大学理学院,江西赣州341000 [2]昆明冶金高等专科学校计算机信息学院,云南昆明650033 [3]昆明冶金高等专科学校人力资源处,云南昆明650033

出　　处：《昆明冶金高等专科学校学报》2015年第3期60-64,共5页Journal of Kunming Metallurgy College

摘　　要：借鉴神经网络里面竞争学习的思路,通过引入一个抑制因子,来提高最大隶属度的值,对应减小其它隶属度的值,以便达到更快的分类速度,同时实现既突出主要因素,又抑制次要因素的目的。提出了基于粗集与FCM的快速图像分割的改进方法。实验结果表明,该快速图像分割方法,对于被噪声污染的图像有较好的分割性能,粗集理论在处理不确定性信息方面有着独特的方式和相关信息提取能力,以及和其他智能方法的易融合性,使得粗糙集理论在图像处理领域有良好的应用前景。In this paper,the competitive learning neural network for reference,through the introduction of an inhibitory factor,to maximize the value of membership,the other corresponding decreases the value of mem-bership,in order to achieve faster classification.As to highlight the main factors and secondary factors,can inhibit the purpose.The improved method of rough set and fast image segmentation based on FCM was put forward.Experimental results show that the fast image segmentation method for image noise pollution has the better segmentation performance,rough set theory has a unique way of dealing with uncertain information, the ability to extract relevant information and easy integration of other intelligent methods.Rough set theory has a good application prospect in the field of image processing.

关 键 词：粗糙集 图像分割 模糊聚类 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]