拉深件缺陷在线检测系统设计  

Design of on-line detection system for drawing defects

作　　者：徐信[1] 肖小亭[1] 温锡昌 谢家静 

机构地区：[1]广东工业大学材料与能源学院,广东广州510006

出　　处：《锻压技术》2016年第7期97-101,共5页Forging & Stamping Technology

基　　金：广东省战略性新兴产业发展专项(2012A090100014);广东工业大学2015年大学生创新项目(yj201511845110)

摘　　要：针对拉深缺陷人工检测的不足问题,提出了一种生产线上表面缺陷检测系统的设计方案。通过Halcon和VC++的联合开发,实现图像处理流程以及检测结果输出,包括摄像机标定、对图像进行平滑处理、利用动态阀值分割表面缺陷特征,通过形态学处理识别破裂缺陷,并对破裂缺陷分类和测量以及输出破裂缺陷外形和尺寸等,实现冲压生产线拉深件表面缺陷的自动检测功能。其中在相机标定过程中,采用Halcon的相机标定方法,快速获取图像坐标与世界三维坐标的对应关系,并实现缺陷特征的准确测量。此方案具有一定的稳定性,可提高检测效率。For the problem of manual detection drawing defects,the design scheme of an on-line automatic detection system was put forward on the stamping production line. In the scheme,the image processing and the output of detection were achieved through the joint development of Halcon and VC ＋ ＋ including camera calibration,image smooth processing,the segment of surface defects with dynamic threshold. Furthermore,the auto-detection function on surface defects of drawing parts in stamping production line was achieved by identification of crack defects with morphology processing,classification and measurement of crack defects and output of the length and the shape of the crack defects etc. During the camera calibration,camera calibration method based on Halcon was carried out,and the transformation between image coordinate and three-dimensional coordinates of the world could be obtained to achieve automatic measurement of defect features accurately. This scheme is of good stability,and the detection efficiency is improved.

关 键 词：冲压生产线 冲压缺陷 图像处理 在线检测 相机标定 

分 类 号：TG385.9[金属学及工艺—金属压力加工]