改进canny算子的亚像素定位算法  

Sub-Pixel Localization Algorithm Based on Modified Canny Operator

作　　者：舒启林[1] 山博 SHU Qi-lin;SHAN Bo(College of Mechanical Engineer,Shenyang Ligong University,Liaoning Shenyang 110159,China)

机构地区：[1]沈阳理工大学机械工程学院,辽宁沈阳110159

出　　处：《机械设计与制造》2018年第10期165-168,共4页Machinery Design & Manufacture

基　　金：辽宁省科技工业攻关项目(2011220009)

摘　　要：为了提高亚像素边缘定位精度,减小定位误差,提出了一种改进的canny边缘检测算子用来检测图像的像素级边缘,之后基于改进的canny算子粗定位的边缘点,采用高斯拟合亚像素方法找出图像的亚像素边缘点位置。针对微小的圆形零件进行图像采集及图像处理,通过matlab实验仿真将该方法与传统的canny算子相比较,发现误定位明显减少,在保留边缘信息的同时有效的提高了边缘定位的精度,得到更准确的圆心位置及半径长度。结果表明是一种有效的边缘检测定位算法,具有一定的实用意义。In order to improve the sub-pixel edge location accuracy and reduce the location error,this paper proposed an improved Canny edge detection operator to detect pixel edge of the image,then on the base found the sub-pixel edge location by Gaussian fitting sub-pixel method.This method collected and processed the image of small circular parts,and compared with the traditional canny operator by the matlab simulation.It was found that the method reduced location error significantly,reserved the edge information and effectively improved the localization accuracy at the same time,Obtained a more accurate center position and length of the radius.The results indicated that his method was an effective edge localization algorithm,and had certain practical significance.

关 键 词：图像处理 边缘检测 CANNY算子 圆心定位 亚像素 

分 类 号：TH16[机械工程—机械制造及自动化] TP751.1[自动化与计算机技术—检测技术与自动化装置]