基于视觉的管道内焊缝定位  

Weld Seam Positioning in Pipeline Based on Vision

作　　者：郝超超 王子文[1] 

机构地区：[1]河北联合大学机械工程学院,河北唐山063009

出　　处：《机械工程师》2015年第9期46-49,共4页Mechanical Engineer

摘　　要：针对管道机器人对燃气管道接头焊缝进行打磨、喷涂防腐材料时需要焊缝坐标信息的问题,文中所研究的定位方法是根据采集的焊缝图像,采用ROI提取与OSTU法结合的分割算法,运用形态学算法去除孤立噪点,通过Sobel算子实现边缘提取,之后运用两次曲线拟合对图像进行拟合,最终从焊缝图像中提取出焊缝位置信息,拟合的焊缝位置与真实焊缝位置吻合程度高。To obtain the welding seam position information of gas pipeline for welding seam polishing and spraying anticorrosion material using pipeline robot, a positioning method is designed. Image segmentation and morphological algorithms which combines ROI（interested image regions） extraction algorithm with OSTU method are adopted to remove isolated noise. Edge detection is realized using Sobel, and eventually weld location information is extracted from welding image using curve fitting, the fitting weld welding seam position is consistent with the real weld welding seam position.

关 键 词：管道 焊缝定位 形态学 

分 类 号：TG441.7[金属学及工艺—焊接] TP212[自动化与计算机技术—检测技术与自动化装置]