基于视觉的大尺寸板材冲孔质量检测系统设计  

Design of Vision⁃based Punching Quality Detection System for Large⁃size Sheet Metal

作　　者：赵海文[1] 郑锦云 张雅丽[1] ZHAO Hai-wen;ZHENG Jin-yun;ZHANG Ya-li(School of Mechanical Engineering,Hebei University of Technology,Tianjin 300401,China)

机构地区：[1]河北工业大学机械工程学院,天津300401

出　　处：《仪表技术与传感器》2020年第11期79-83,共5页Instrument Technique and Sensor

摘　　要：针对自动化产线中板材冲孔质量多为人工抽检问题,设计了一套基于视觉的大尺寸板材冲孔质量自动检测系统。该系统利用二维视觉检测平台对板材进行图像采集,通过中值滤波、二值化分割和Canny边缘检测算子等算法对图像进行处理;利用坐标变换模型计算不同图像中孔的相对位置关系,进而确定所有孔相对于板材坐标原点的位置。最后通过实验,将视觉检测结果与人工测量数据进行对比,验证了该系统的可行性及准确性。In view of the fact that the punching quality of sheet metal in automatic production line is mostly manually sam⁃pled,an automatic detection system for punching quality of large⁃size sheet metal based on vision was designed,which used two⁃dimensional vision detection platform to collect images of sheet metal,and then processed the collected images by algorithms such as median filtering,binary segmentation and canny edge detection operator and so on.According to the coordinate transformation model,the relative positions of holes in different images were computed,and then the positions of all holes relative to the origin of plate coordinates were determined.Finally,the feasibility and accuracy of the detection system was verified by comparing the visual detection results with the manual measurement data.

关 键 词：大尺寸板材 冲孔质量 视觉检测 图像处理 坐标变换 检测平台 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]