袋式输液器液袋组件管路缺失视觉检测研究  

Study on Visual Detection of Missing Pipeline of Liquid Bag Assembly of Bag Infusion Device

作　　者：黄波 刘香 张骞 刘康 刘建宏 HUANG Bo;LIU Xiang;ZHANG Qian;LIU Kang;LIU Jianhong(School of Mechanical Engineering,Sichuan University of Science&Engineering,Yibin 644000,China)

机构地区：[1]四川轻化工大学机械工程学院,四川宜宾644000

出　　处：《四川轻化工大学学报（自然科学版）》2024年第3期34-41,共8页Journal of Sichuan University of Science & Engineering(Natural Science Edition)

基　　金：四川省科学技术厅项目(2021YFG0050);人工智能四川省重点实验室项目(2020RYY01)。

摘　　要：人工质检液袋组件会因主观性和疲劳带来质检风险,还会降低生产线的自动化程度。为解决这一问题,采用机器视觉系统对液袋组件挂绳、加药两通进行组装质量检测。设计图像采集辅助机构,解决因液袋组件形态不固定而难以进行图像采集的难题;在确定缺陷检测策略的基础上,选用Canny算法以最短时间提取液袋管路区域的清晰轮廓,相较于其他算法节约160~520ms;采用虚拟直线法,结合本文根据轮廓外接矩形框位置特征选定的感兴趣区域(ROI),完成轮廓数的统计,最终实现缺陷检测目标。实验表明,管路缺失检出率达100%,检测最高耗时438ms,检测一组液袋组件仅需要1.75s,远小于工厂生产线上生产一组液袋组件的单工位工时8s。该方法能代替现有人工肉眼检测,减少企业用人成本。Manual quality inspection of the liquid bag assembly will bring quality inspection risk due to subjectivity and fatigue and also reduce the automation degree of the production line.To solve this problem,the machine vision system is adopted to inspect the assembly quality of the lanyard and dosing two passes of liquid bag components.The image acquisition assisting mechanism is designed to solve the problem of difficult image acquisition due to the unfixed shape of the liquid bag assembly;based on determining the defect detection strategy,the Canny algorithm is selected to extract the clear contour of the liquid bag pipeline area in the shortest time,which saves 160~520 ms compared with other algorithms;using the virtual straight-line method,combined with the region of interest(ROI)selected by this paper based on the location characteristics of the outer rectangular box of the contour,the statistics of the number of contours are completed,and the defect detection goal is finally realized.Experiments show that the pipeline defect detection rate of 100%,the maximum detection time of 438 ms,and the detection of a group of liquid bag components need only 1.75 s,much smaller than the factory production line production of a group of liquid bag components of a single station working time of 8 s.The method can replace the existing manual visual inspection,which reduces the cost of enterprise employment.

关 键 词：机器视觉 袋式输液器 缺陷检测 图像处理 边缘检测 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]