基于图像过渡区的CO_2气体保护焊焊缝识别算法  

A CO_2 arc welding seam detection algorithm based on transition region

作　　者：杨福刚[1] 孙同景[1] 张光先[1] 庞清乐[1] 

机构地区：[1]山东大学控制科学与工程学院,济南250061

出　　处：《焊接学报》2007年第2期89-91,共3页Transactions of The China Welding Institution

摘　　要：利用过渡区图像分割技术实现CO2气体保护焊焊缝的识别。采集到的焊缝图像在目标区和背景区存在过渡区域,在直方图上表现为双峰之间的谷比较宽广且平坦。通过对原始图像进行高端剪切和低端剪切变换,利用不同灰度像素的梯度信息来确定过渡区,进而利用过渡区来确定图像分割的阈值。克服了图像中非目标区域的影响,可以抗噪声和干扰。算法简单,实时跟踪性也较好。结果表明,该方法边缘检测效果良好,是一种比较适合CO2气体保护焊焊缝识别的方法。In order to detect CO2 arc welding seam,a new image segmentation method based on transition region is presented.Transition region is a special region located between the object and background in the real image,whose histogram has a wide and evenly hollow between two peaks.First,transition region of original image is determined by calculating the average grads of nonzero pixels of the image obtained through top-cut and bottom-cut transform,and then threshold can be obtained easily from transition region.It overcomes the effect of disturbance.Examinations indicate that it is a good method to detect CO2 arc welding seam.

关 键 词：CO2气体保护焊 边缘检测 焊缝识别 过渡区 

分 类 号：TG403[金属学及工艺—焊接]