激光视觉传感器对不连续焊缝的识别  

Identification of discontinuous welding seam using laser vision sensor

作　　者：王艳庆[1] 程伟德[1] 张华[1] 刘冬花[1] 

机构地区：[1]南昌大学信息工程学院机器人研究所,江西南昌330031

出　　处：《传感器与微系统》2011年第5期8-11,15,共5页Transducer and Microsystem Technologies

基　　金：国家"863"计划资助项目(2007AA04Z242)

摘　　要：船舶制造中,为了对小格子船舱中排水孔形成的不连续焊缝进行跟踪,在机器人上安装视觉传感器提取排水孔的起点、终点特征信号。介绍了图像的采集方法和图像处理过程。对含有焊缝信息的图片先进行去噪预处理,再进行阈值分割,求取连通区与面积;细化图像,提取出激光线条的骨架;然后再进行霍夫变换,提取特征直线;最后求得两直线的交点,即焊缝的特征点,从而识别排水孔和焊缝。在模拟船舱结构的工件上进行不连续焊缝跟踪实验,实现了不连续焊缝的识别。In order to identify the discontinuous welding seam created by drainage holes in lattice-type cabin in shipbuilding,a vision sensor is set up in the robot to extract the startpoint and the endpoint of drainage holes.The way of image acquisition and image processing procedure are introduced.The image of welding seam is denoised.Threshold segmentation are done and connected region and their area are striked.The image is refined and the frame of laser line is extracted.Characteristic line is extracted using Hough transform.The intersection of two lines is striked,which is the characteristic point of welding seam.So that drainage holes and welding seam are identified.Discontinuous welding seam tracking experiment on workpiece that similar to cabin are carried out,and the experiment is successful.

关 键 词：图像处理 不连续焊缝 焊接自动化 视觉传感器 

分 类 号：TP216[自动化与计算机技术—检测技术与自动化装置]