一种基于改进的SUSAN算法的焊缝边缘检测方法  

A Weld Edge Detection Mehod Based on Improved SUSAN Algorithm

作　　者：陈石涛[1] 杨龙兴[1] 丁力[1] 梁栋[1] CHEN Shitao;YANG Longxing;DING Li;LIANG Dong(Colledge of Mechanics,Jiangsu University of Technology,Changzhou 213001,China)

机构地区：[1]江苏理工学院机械学院

出　　处：《热加工工艺》2018年第15期161-164,共4页Hot Working Technology

基　　金：江苏省基础研究计划(自然科学基金)资助项目(BK20170135)

摘　　要：随着图像技术在自动化焊接中的应用,高效地识别和提取出焊缝轨迹,成为图像跟踪自动焊接的关键。本文提出了一种基于改进SUSAN算法的焊缝边缘检测方法。首先对采集到的焊缝图像进行灰度化处理,再通过SUSAN算法进行焊缝的边缘检测,并将SUSAN算法中的灰度差阈值由人为选取改进为一种自适应选取的方法。理论分析和实验结果表明,该算法比其他边缘检测算子抗噪能力强,可以更加高效地提取出焊缝信息。With the application of image technology in automatic welding, it is the key to realize the automatic welding of image tracking to efficiently identify and extract the weld track. A weld edge detection method based on improved SUSAN algorithm was presented. Firstly, the collected welding images were processed by grayscale, the weld edge was detected through SUSAN algorithm, and the gray level difference threshold in SUSAN algorithm was improved from human selection to an adaptive selection method. The theoretical analysis and experimental results show that the algorithm has stronger anti-noise property and can extract the weld information more efficiently than other edge detection algorithms.

关 键 词：焊缝图像 SUSAN算法 焊缝边缘检测 灰度化 自适应选取 

分 类 号：TG409[金属学及工艺—焊接]