摘要
边缘检测是提取图像特征的首要条件。文章提出了一种改进型Canny边缘检测算法。首先在边缘梯度计算时,将二维滤波模板分解为两个一维滤波模板,实现并行处理,提高运算速度;引入双阈值法则同时保证图像中强边缘点和弱边缘点的提取,并基于非局部最大值抑制原理检测边缘点,大大提高了边缘检测的精度和准确度;最后采用形态学算子实现对检测边缘进行细化处理,实现了单像素级细化边缘提取。实验结果表明,该算法在保证实时性的同时,具有很好的检测精度和准确度。
Image edge detection is the first step to obtain image feature.In this thesis,an improved Canny edge detec-tion algorithm is represented to obtain thin and robust edges.Compared with ordinary Canny method,there are four im-provements to reduce computation time and ensure detection accuracy.Firstly,2-D Gaussian filter is decomposed into two independent 1-D filters,i.e.row filter and column filter,which allows calculate image gradient in parallel way.As a result,computation time is reduced highly.Secondly,the method uses two thresholds,to detect strong and weak edges,and includes the weak edges in the output only if they are connected to strong edges.This method is therefore less likely than the others to be'fooled'by noise,and more likely to detect true weak edges.Thirdly,non-maximum suppression principle is adopted to detect true edges.Finally,edge thin operation is conducted based on morphological operator to obtain single pixel level edge.The effectiveness of the proposed method is demonstrated through practical experiment.
出处
《计算机工程与应用》
CSCD
北大核心
2004年第20期27-29,共3页
Computer Engineering and Applications
基金
国家自然科学基金项目(编号:60375001)
关键词
边缘检侧
CANNY算子
边缘细化
自适应阈值
非极值抑制
edge detection,Canny operator,edge thinning,self-adaptive threshold,non-maximal suppression