基于机器视觉的药柱端面银丝检测系统设计

Design of Silver Wire Detection System for Grain End Based on Machine Vision

针对内埋金属丝药柱端面银丝识别过程中人工检测效率低、精度差的问题,设计了一种基于机器视觉的药柱端面银丝检测系统。该系统通过工业相机采集药柱端面图像,使用C++和Opencv库实现银丝检测算法,采用MFC的多线程技术提高算法的运行效率,通过TCP/IPv4协议与可编程逻辑控制器通信。同时,针对切断药柱时银丝被抽出,无法检测银丝位置的问题,提出了一种多点预测模型,提高了银丝检测的稳定性。试验结果表明系统设计应用合理、运行稳定、参数配置简便,银丝识别率可达99.9%,检测速度可满足工业自动化生产的要求。

Aiming at the problems of low efficiency and poor accuracy in manual detection of silver wire in the identification process of embedded metal wire grain end,a silver wire detection system for grain end based on machine vision is designed.The system collects the image of grain end by the industrial camera,uses C++and Opencv library to realize the silver wire detection algorithm,uses the multi-threading technology of MFC to improve the operation efficiency of the algorithm,and communicates with the programmable logic controller through TCP/IPv4.At the same time,aiming at the problem that the silver wire could not be detected when cutting off the column,a multi-point prediction model is proposed to improve the stability of silver wire detection.The tests results show that the system design and application is reasonable,the operation is stable,the parameter configuration is simple,the silver wire recognition rate can reach 99.9%,the detection speed can meet the requirements of industrial automation production.