数字图像处理在焊缝识别中的应用

Application of digital image processing in the weld seam recognition

运用数字图像处理的方法,对电荷耦合器(CCD)摄像头在自然光下拍摄的焊接工件图像进行焊缝特征提取。针对图像的实际特征,采用中值滤波-Laplacian图像锐化-Roberts检测-焊缝提取和中值滤波-自适应阈值处理-焊缝提取两套方案对图像进行处理,提取出焊缝特征。通过实际处理结果的对比,后者能较好地对焊缝进行识别。

This paper applies the digital image processing to the welding workpiece picture, which is shot by CCD video camera in natural light, and detects the weld seam characteristics. According to the weld seam characteristics, two methods are used to detect the weld seam. One is the detection of median filter - Laplacian image sharpen - Roberts operator - abstracting weld seam, the other is the detection of median filter - self- adapt threshold image segmentation - abstracting weld seam. The comparison of the experiment results shows that the later method is a useful way to recognize the weld seam.