二维斜分指数交叉熵的印章图像阈值分割  

Exponential cross entropy thresholding for seal image based on 2-dimensional oblique segmentation

作　　者：万水龙 刘进[1] 余彪[3] 

机构地区：[1]中国人民解放军91960部队 [2]南京航空航天大学 [3]中国人民解放军92493部队

出　　处：《微型机与应用》2013年第24期46-49,共4页Microcomputer & Its Applications

基　　金：文件检验鉴定公安部重点实验室开放课题(10KFKT005)

摘　　要：提出了一种二维斜分指数交叉熵的印章图像阈值分割方法,借鉴二维直方图的斜分思想对二维指数交叉熵直方图进行斜分。实验结果表明,相对于最大熵法和指数熵法,提出的二维斜分指数交叉熵的印章图像阈值分割方法分割效果更佳。The exponential cross entropy thresholding method based on 2-dimensional histogram oblique segmentation is pro-posed by using 2-dimensional histogram vertical idea . A large number of experimental results show that in contrast with the maxi-mum entropy method and the exponential entropy method , the exponential cross entropy thresholding for seal image based on 2-di-mensional histogram oblique segmentation can achieve superior segmented results and greatly reduce the running time .

关 键 词：印章图像 阈值分割 指数交叉熵 斜分 

分 类 号：TN911.73[电子电信—通信与信息系统]