基于Canny算子字符边缘检测与分割方法研究  

Characters Edge Detection and Segmentation Method Based on Canny Operator

作　　者：李国平[1] 路长厚[2] 李健美[2] 

机构地区：[1]济南大学机械学院,济南250022 [2]山东大学机械学院,济南250061

出　　处：《机床与液压》2007年第12期42-44,15,共4页Machine Tool & Hydraulics

基　　金：济南大学校基金(Y0420);教育部博士点基金(20060422011)

摘　　要：字符边缘是标牌图像中的凹凸字符最主要特征,选用适当的边缘检测算子得到准确的标牌信息是字符图像处理的关键步骤。本文叙述Canny算子的基本原理和算法实现过程,选用并研究Canny算子来提取字符的边缘特征。为了验证Canny边缘检测算子的效果,分别用Canny、Robert、Sobel、Prewitt、Laplacian和LoG算子对字符图像进行边缘检测试验,分析了边缘检测的结果,得出Canny边缘检测算子是字符边缘检测的一种十分有效的方法。依据Canny算子边缘检测得到的结果,对字符图像进行了填充并分割得到了标牌字符的分割二值化图像。边缘处理和字符分割的结果表明,Canny算子边缘检测能够较好地解决低对比度图像和灰度不均匀字符图像的二值化问题。Characters edge is the primary feature of protuberant characters on label image. Acquisition accuracy label information through selection edge detection operator is the key step of character image processing. The Canny operator principle and algorithm were described and were used to detect the edges of protuberant characters. In contrast to test the effect of Canny operator, Canny operator, Robert operator, Sobel operator, Prewitt operator, Laplacian operator and LoG operator were taken to detect character edge. The resuits of edge detection show that Canny operator is most fit to detect the characters edge. According to the results of edge detection with Canny operator, the edge detection character image was filled and the character image was segmented. The experiments of edge detection and character segmentation prove that Canny operator can preferable solve the binarization problems of low contrast image and uniform gray character image.

关 键 词：CANNY算子 边缘检测 字符分割 图像处理 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]