基于计算机视觉的无人值守过磅系统设计  

Design of Computer Vision-Based Unattended Weighing Scale System

作　　者：马世欢[1] 李伟[1] MA Shihuan;LI Wei(School of Electronic and Information Engineering,Henan Polytechnic Institute,Nanyang 473000,China)

机构地区：[1]河南工业职业技术学院电子信息工程学院,河南南阳473000

出　　处：《自动化仪表》2022年第5期12-14,19,共4页Process Automation Instrumentation

基　　金：2019年河南省高等教育教学改革研究与实践基金资助项目(2019SJGLX701)。

摘　　要：无人值守过磅系统是现代化工矿企业提高工作效率的重要手段。在对现有系统进行研究的基础上,设计了基于计算机视觉的无人值守过磅系统。该系统可通过计算机视觉技术,进行车牌识别;通过目标检测和跟踪算法,进行车辆运行状态和位置的判断;通过物联网(IoT)技术,进行数据和控制指令的传输。实际应用证明,该系统计量误差在3‰以内,符合工业计量要求。系统采用嵌入式设备,节省了安装空间;使用计算机视觉技术判断车辆状态,减少了传感器的数量;使用IoT技术,提高了系统的可扩展性,降低了布线工作量。这些特点简化了系统的部署和维护,降低了成本,具有较高的应用价值。该系统的使用对于计算机视觉技术在工业领域的应用具有积极意义。Unattended weighing scale system is an important means to improve the efficiency of modern industrial and mining enterprises.Based on the study of the existing system,a computer vision-based unattended weighing scale system is designed.The system can carry out license plate recognition by computer vision technology,judgment of vehicle running status and position by target detection and tracking algorithm,and transmission of data and control commands by Internet of Things(IoT)technology.The actual application verifies that the system's measurement error is within 3‰,which meets the industrial measurement requirements.The system uses embedded devices to save installation space,uses computer vision technology to determine vehicle status,reduces the number of sensors,and uses IoT technology to improve the scalability of the system and reduce wiring workload.These features simplify the deployment and maintenance of the system,reduce the cost,and have high application value.The use of this system has positive implications for the application of computer vision technology in the industrial field.

关 键 词：过磅系统 目标检测 对象跟踪 计算机视觉 物联网 树莓派 ESP8266 MicroPython 

分 类 号：TH-39[机械工程]