基于局部熵的边缘检测算子选择算法  

Algorithm of edge detector choice based on local entropy

作　　者：凌军[1] 宋启祥[1] 房爱东[1] 李现伟[1] Ling Jun;Song Qixiang;Fang Aidong;Li Xianwei(Information Engineering College,Suzhou University,Suzhou 234000,China)

机构地区：[1]宿州学院信息工程学院,安徽宿州234000

出　　处：《南京理工大学学报》2018年第4期424-429,共6页Journal of Nanjing University of Science and Technology

基　　金：宿州区域发展协同创新中心课题资助(2016szxt06;2016szxt05);宿州学院2015年质量工程项目(szxy2015jy12);安徽省高等学校质量工程研究重大项目(2016jyxm1026)

摘　　要：为了在图像处理中选择恰当的边缘检测算子,保证边缘提取的质量,提出边缘点加权平均边缘局部熵,对利用边缘点加权平均边缘局部熵选择边缘检测算子的算法进行了研究。通过边缘检测算子抽取出图像的边缘,计算图像边缘的局部熵,使用边缘像素的灰度信息进行加权,计算出边缘点加权平均边缘局部熵。依据边缘点加权平均边缘局部熵进行边缘检测算子的选择。实验结果表明:算法能有效地选择合适的边缘检测算子,保证获取的图像边缘质量。To choose suitable edge detectors in image processing and promise quality of edge extraction,an edge-point-weight average edge local entropy is presented.A choice algorithm of edge detectors with edge-point-weight average edge local entropy is studied.Firstly,image edge is drawn out by edge detector and local entropy of image edge is caculated.Secondly,edge-point-weight average edge local entropy of image edge is obtained by weighting local entropy of image edge with edge grey.Edge detector is chosen with edge-point-weight average edge local entropy.Experimental result demonstrates that the algorithm is valid in choosing suitable edge detector and can promise quality of image edge.

关 键 词：算子选择 边缘检测 边缘点加权 局部熵 边缘局部熵 边缘点加权平均边缘局部熵 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]