基于OpenCV的搭接型焊缝特征点提取控制工程专业论文.docxVIP

山东科技大学工程硕士学位论文摘要 山东科技大学工程硕士学位论文 摘要 摘要 随着焊接自动化水平的日益提高，焊缝跟踪系统成为焊接领域的一项重要的研究课 题。焊缝特征点像素坐标的准确度直接影响着焊缝跟踪系统的精度，该问题的研究具有 重要的理论和实际意义。 本文主要研究利用激光条纹辅助焊缝特征点的提取，为机器人自动焊接的实现提供 有效的帮助。首先，本文针对机器人末端安装的双目摄像头开发了焊缝图像采集系统； 其次，对激光束打在工件上的激光条纹图像进行预处理，例如灰度化、开运算、中值滤 波、二值化、细化等等，得到单像素的焊缝图像；最后采用了直线拟合法、曲率法、斜 率法、Hough 变换法分别提取焊缝特征点的像素坐标，并且对四种提取焊缝特征点像素 坐标的方法的性能进行了对比，选用 Hough 变换方法作为本文提取焊缝特征点的方法。 本论文利用 MOTOMAN 机器人、Bumblebee2 双目工业摄像头、1394 图像采集卡等 硬件设备和 Visual Studio2008 集成开发环境、OpenCV 等相关软件搭建了研发平台。并 经过图像预处理和焊缝特征点提取算法实现了对焊缝特征点像素坐标的提取，完成了具 有焊缝特征点像素坐标提取功能的软件开发工作。为以后利用 MOTOMAN 机器人实现 焊接自动化打下了良好基础。 关键词：双目相机，激光条纹提取，焊缝特征点提取，Hough 变换 ABSTRACT With the increasing of the welding automation level, welding seam tracking system becomes an important research topic in the field of welding. The accuracy of the pixel coordinates of the feature points of weld has a direct impact on the accuracy of the seam tracking system, the issue has important theoretical and practical significance. This paper mainly studies the extraction of weld feature point which is auxiliary by laser stripe, and it provides effective help for the realization of robot automatic welding. Firstly, this paper developed weld image acquisition system that is based on binocular camera which is installed in the end of the robot. Secondly, this paper preprocessed the laser stripe image which is formed by laser beaming on the work pieces, for example, image gray processing, opening operation, median filter, binary image and thinning, to obtain weld image of a signal pixel. At last this paper used algorithms of linear fitting, curvature, slope and Hough transformation to extract the pixel coordinates of the weld characteristic point and comparing the quality of the four methods, this paper decided to use Hough transform to extract pixel coordinate of weld seam characteristic point. By using the hardware device of MOTOMAN robot of Japan, the Bumblebee2 industrial camera of two eyes and the video capture card of 1394 and the related software of Microsoft’s Vis