基于结构光视觉的不锈钢角焊缝检测关键技术研究

发布时间：2018-05-20 16:29

本文选题：结构光视觉 + 焊缝检测　；参考：《深圳大学》2017年硕士论文

【摘要】：焊接凭借其可靠性强、精确度高、成本低等优点成为现代工业生产的重要金属连接加工技术。然而人工焊接难以满足日益提高的焊接速度和焊接质量的要求,且焊接过程的弧光辐射和烟尘等给长期从事焊接的工人身体带来危害。利用机器人进行自动化焊接来取代人工焊接是解决上述问题的方法之一,也是焊接加工行业的发展趋势。焊缝检测技术是实现机器人自动化焊接的关键技术,其中基于结构光视觉的焊缝检测技术具有精度高、速度快、非接触等优点,具有重要的研究与应用价值,本文对基于结构光视觉的不锈钢角焊缝检测关键技术进行研究。本文综合研究了焊缝检测技术的分类与发展、基于结构光视觉的焊缝检测技术的国内外研究现状,针对不锈钢角焊缝的检测提出技术方案。研究了摄像机模型、基于结构光视觉的三角测量原理并设计搭建了基于CMOS型工业摄像机、镜头、滤光片、结构光发生器的图像采集终端,根据工程需求论证确定了该图像采集终端与机器人间的Eye-in-Hand装配方式。研究了摄像机参数、结构光平面空间位置参数以及基于结构光视觉的图像采集终端和机器人间装配参数的标定理论与方法,采用Matlab软件的标定工具箱以及自主编写的算法程序进行实验,计算并优化得到上述参数。对不锈钢角焊缝图像特征进行了分析,提出了以Region of Interest(ROI)选定、图像滤波、特征检测、特征提取为主要流程的图像处理算法,并基于OpenCV库函数进行编程实现。通过对提取的不锈钢角焊缝特征点的像素坐标经过坐标转换得到对应机器人基坐标系下的焊缝特征点空间位置坐标。本文对所研究的基于结构光视觉的不锈钢角焊缝检测关键技术进行了实验验证。实验表明所设计搭建的基于结构光视觉的图像采集终端实现了测量功能,测量误差小于±0.145mm,平均误差为0.090mm;最终实现了焊缝位置的检测,焊缝位置检测与参考位置最大偏差小于0.96mm,平均偏差为0.35mm。
[Abstract]:Welding has become an important metal connection processing technology in modern industry because of its high reliability, high accuracy and low cost. However, manual welding is difficult to meet the requirements of increasing welding speed and welding quality, and the arc radiation and smoke in the welding process are harmful to the health of workers engaged in welding for a long time. Using robot to replace manual welding is one of the methods to solve the above problems, and it is also the development trend of welding processing industry. Welding seam detection technology is the key technology to realize robot automatic welding. The welding seam detection technology based on structured light vision has the advantages of high precision, high speed, non-contact and so on, so it has important research and application value. In this paper, the key technology of stainless steel fillet seam detection based on structured light vision is studied. In this paper, the classification and development of welding seam detection technology are comprehensively studied. The research status of welding seam detection technology based on structured light vision at home and abroad is comprehensively studied, and the technical scheme for the detection of stainless steel fillet weld is put forward. The camera model, triangulation principle based on structured light vision and the image acquisition terminal based on CMOS industrial camera, lens, filter and structured light generator are designed and built. The Eye-in-Hand assembly mode between the image acquisition terminal and the robot is determined according to the engineering requirement demonstration. The calibration theory and method of camera parameters, space position parameters of structured light plane and assembly parameters of image acquisition terminal and robot based on structured light vision are studied. The calibration toolbox of Matlab software and the algorithm program written by ourselves are used to carry out experiments, and the above parameters are calculated and optimized. The image features of stainless steel fillet weld are analyzed, and an image processing algorithm based on Region of InterestROI, image filtering, feature detection and feature extraction is proposed. The algorithm is programmed based on OpenCV library function. Through the transformation of the pixel coordinates of the stainless steel corner weld feature points, the corresponding space coordinates of the weld feature points in the robot base coordinate system are obtained. In this paper, the key techniques of stainless steel fillet seam detection based on structured light vision are experimentally verified. The experimental results show that the designed image acquisition terminal based on structured light vision realizes the measurement function, the measuring error is less than 卤0.145 mm and the average error is 0.090 mm. The maximum deviation between weld position detection and reference position is less than 0.96mm, and the average deviation is 0.35mm.
【学位授予单位】：深圳大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TG441.7;TP391.41

【相似文献】