现代电子技术2024，Vol.47Issue(23)：36-42,7.DOI:10.16652/j.issn.1004-373x.2024.23.006
一种改进Canny算子的图像边缘检测方法
Image edge detection method based on improved Canny operator
摘要
Abstract
The traditional Canny edge detection algorithm fails to deal with salt-and-pepper noise interference,and the threshold selection depends on manual intervention,so an edge detection algorithm based on improved Canny operator is proposed.An improved recursive median filtering is proposed to replace the Gaussian filter,which can reduce the loss of image edge information while removing image noise.Then bilinear interpolation is used to improve non-maximum suppression to locate the pixel gradient values at the level of sub-pixel.The iteration-based Otsu threshold segmentation algorithm is used to obtain the threshold value adaptively and perform the iterative edge connection,so as to improve the adaptive degree of the algorithm.Experimental results show that,in comparison with the algorithm based on Canny operator and the algorithm based on Sobel operator,the denoised image quality evaluation index of the proposed algorithm is improved by 8.5%,and its edge detection accuracy is improved by 22%.Moreover,the proposed algorithm has outstanding ability of removing salt-and-pepper noise,strong adaptability,more complete edge information detected and better continuity.关键词
图像处理/Canny/边缘检测/改进的递归中值滤波/双线性插值/基于迭代的Otsu阈值分割Key words
image processing/Canny/edge detection/improved recursive median filtering/bilinear interpolation/iteration-based Otsu threshold segmentation分类
信息技术与安全科学引用本文复制引用
包启龙,景会成,张靖轩,王一,杨富荃..一种改进Canny算子的图像边缘检测方法[J].现代电子技术,2024,47(23):36-42,7.基金项目
河北省教育厅科学研究项目资助(CXY2024013) （CXY2024013）
河北省省级科技计划资助(21314303D) （21314303D）
唐山市科技计划项目(21130219C) （21130219C）
