一种基于粗糙集的脉冲噪声图像边缘检测算法  

A kind of based on rough set of impulse noise image edge detection algorithm

作　　者：刘勍[1] 魏弘博 应隽[3] 赵小龙[1] 

机构地区：[1]天水师范学院物理与信息科学学院,甘肃天水741001 [2]陕西省军区司令部通信处,陕西西安710061 [3]合肥工业大学计算机与信息学院,安徽合肥230009

出　　处：《自动化与仪器仪表》2011年第6期123-124,127,共3页Automation & Instrumentation

基　　金：甘肃省自然科学基金计划项目(1010RJZE028);天水师范学院中青年教师科研项目(TSA1001)及天水师范学院"青蓝"人才工程基金项目资助

摘　　要：研究并提出了一种图像边缘检测算法,对被脉冲噪声污染的图像进行预处理,从而便于图像的分析、理解等后续操作。与一般算法不同,该算法从图像的边缘特性入手来解决边缘检测问题,以由特性决定的一系列边缘点的约束条件作为算法基础,并用粗糙集理论解决了这些条件的相关问题,进而建立了整个边缘检测算法。计算机仿真表明,本文的算法能够有效地从含噪图像中提取边缘信息,较好地克服了传统算法对噪声的敏感性问题。A novel edge-detecting algorithm is proposed to process picture that is polluted by salt and pepper noise in this paper.Different from other approaches,it starts from the characteristics of image edges.As a theory foundation,a serial of edge point constraints are defined first,and then it solves these constraints with the knowledge of rough sets theory.As the foundation is constructed,the new algorithm is put forward successfully.Experimental results show that this new algorithm works effectively and produces good edges.Simultaneity,simulations show that the new algorithm solves the sensitive problem of traditional algorithm.

关 键 词：粗糙集 脉冲噪声 灰度图像 边缘检测 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]