基于同位深度-灰度图像的焊缝成形质量检测  

Weld Forming Quality Detection Based on Depth and Gray Images at Same Time and Same Position

作　　者：闫中广 闫志鸿[2] 谷文涵 YAN Zhongguang;YAN Zhihong;GU Wenhan(Yangquan Quality Technical Supervision Inspection and Testing Institute,Yangquan 045000,China;Engineering Research Center for Advanced Manufacturing Technology of Automotive Structural Components,Faculty of Materials and Manufacturing,Beijing University of Technology,Beijing 100124,China)

机构地区：[1]阳泉市质量技术监督检验测试所,山西阳泉045000 [2]北京工业大学材料与制造学部,汽车结构部件先进制造技术教育部工程研究中心,北京100124

出　　处：《热加工工艺》2021年第15期102-105,111,共5页Hot Working Technology

基　　金：国家自然科学基金面上项目(51975015)。

摘　　要：随着焊接自动化技术的普遍应用,焊缝成形和焊接外观缺陷的自动化检测已成为一种迫切的需求。采用激光结构光视觉检测方法,同位获取焊缝的深度-灰度图像。分析了铝合金焊缝各种缺陷在深度图像和灰度图像中的表现。结果表明,深度图像有利于成形缺陷的检测,而灰度图像更有利于表面裂纹的检查。开发了相应的缺陷检测算法,实现了以上缺陷的提取。With the widespread application of welding automation technology, automatic detection of weld forming and welding surface defects has become an urgent demand. The laser structured light vision method was used to obtain the depth and gray images of weld bead at the same time and same position. The performance of various defects of aluminum alloy weld in depth image and gray image was analyzed. The results show that the depth image is beneficial to the detection of forming defects, and the gray image is more beneficial to the detection of surface cracks. And the corresponding defect detection algorithm was developed to realize the extraction of the above defects.

关 键 词：焊缝成形质量 激光结构光视觉 深度-灰度图像 缺陷提取 

分 类 号：TG409[金属学及工艺—焊接]