基于图像过渡区的CO2气体保护焊焊缝识别算法
A CO2 arc welding seam detection algorithmbased on transition region
-
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。Abstract: 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.