摘要
针对现有水表检定装置检定效率低、信息管理功能不足等问题,对水表指针识别、消除气泡影响、检定台自动化改造等方面进行了研究,提出了一种基于机器视觉的水表在线检测系统。采用距离法读取水表指针读数,通过模板匹配,旋转校正水表图像,利用极坐标变换,将子表盘展开成矩形,并集成二维码识别模块,实现了对水表的检定与数据保存;设计了上行走机构带动相机移动拍照,使用单个相机完成了5个工位水表图像的采集。研究结果表明:与人工肉眼读表和手工记录读数方法相比,该检测方法能确保水表的检定精度,同时提高了检定效率,并且检定数据可溯源。
Aiming at the problems of low calibration efficiency and insufficient information management function of existing water meter calibration devices, the identification of water meter pointer, eliminating the influence of bubbles, and the automatic transformation of calibration platform were studied. A kind of on-line water meter detection system based on machine vision was proposed, distance method was used to read water meter pointer reading and passed through template matching, the water meter image were rotated and corrected, polar coordinate transformation was used to expand the sub-dial into a rectangle, and two-dimensional code recognition module were integrated, the verification and data preservation of the water meter was realized. A walking mechanism was designed to drive the camera to take pictures, and a single camera was used to collect images of water meters at five stations.The results indicate that compared with manual reading and manual recording, this method can ensure the accuracy of water meter, improve the efficiency of verification, and make the calibration data traceable.
作者
郑文学
岳智珉
唐旭晟
陈丹
ZHENG Wen-xue;YUE Zhi-min;TANG Xu-sheng;CHEN Dan(College of Mechanical Engineering and Automation,Fuzhou University,Fuzhou 350108,China;Zhengzhou KaiYuan Instrument Company Limited,Zhengzhou 450000,China)
出处
《机电工程》
CAS
北大核心
2019年第3期271-274,共4页
Journal of Mechanical & Electrical Engineering
基金
福建省自然科学基金资助项目(2018J01534)
关键词
距离法
水表检定
机器视觉
distance method
water meter
machine vision
