基于机器视觉的复杂平面零件尺寸精密检测  

Size Sophisticated Detection Based on Machine Vision Complex Plane Parts

作　　者：付泰[1] 王桂棠[1] 程书豪 熊峰[1] 

机构地区：[1]广东工业大学信息工程学院,广东广州510006

出　　处：《机电工程技术》2016年第8期7-9,84,共4页Mechanical & Electrical Engineering Technology

基　　金：广东省科技计划项目(编号:2015A030401088)

摘　　要：针对精冲汽车零件等尺寸变化较大、局部尺寸精度要求高的零件,采用机器视觉技术进行在线精密尺寸测量。从CAD图纸文件中提取零件形状尺寸和公差等参数,生成对应的运动控制参数,实现相机沿复杂零件轮廓摄取局部图像以提高分辨率,同时运动控制系统反馈编码器得到相机的运动坐标参数。综合两者可得高精度的检测参数,再对比公差要求,即实现零件的合格判断和分类。实验表明对精冲零件的测量精度能达到0.01 mm。该方法为多品种、少批量的平面复杂零件自动测量提供了方便、可靠的解决方案。Fine blanking car parts for dimensional change is large, the size of the local high precision parts using machine vision technology for online precision dimensional measurement. Extraction part shape dimensions and tolerances and other parameters from CAD drawing files, generate a corresponding motion control parameters to achieve the camera along the contour of complex parts in order to improve the resolution of the image pickup locally, while the motion control system to give feedback encoder camera motion coordinate parameters. Combination of the two available high-precision testing parameters, then compare the tolerance requirements , namely to achieve a qualified judge and classify parts. Experiments show that precision fine blanking parts can reach 0.01mm. This method is more variety, less bulk planar complex parts automatic measurements provide a convenient, reliable solution.

关 键 词：CAD 图纸参数 机器视觉 自动测量 平面零件 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]