基于机器视觉的指针式仪表检测方法  

Method for Testing of Pointer Instrument Based on Machine Vision

作　　者：陈树[1] 王磊[1] 

机构地区：[1]江南大学物联网工程学院,无锡214000

出　　处：《计算机与数字工程》2016年第9期1821-1826,共6页Computer & Digital Engineering

基　　金：江苏省六大人才高峰基金(编号:2012-WLW-006)资助

摘　　要：针对传统指针式仪表耐久检测过程中存在的效率低,劳动强度大,精度不高等问题,论文把机器视觉和仪表检测两者融合,通过视觉实时图像的二值化、刻度轮廓提取、拟合圆、指针细化、求解指针直线方程、刻度圆环全景展开等步骤,实时读取表盘数值。实验结果表明:该算法具有鲁棒性好,准确率高,实时性强等优点。Focusing on the problem of low efficiency,high labor intensity,low accuracy in traditional pointer instrument durability test.In this paper,the machine vision and instrument detection are combined.Through the image of binarization,scale contour extraction,fitting circle,pointer refinement,the solution of the linear equation of the pointer,panoramic expansion of the scale ring and other steps,read the dial value in real time.Experimental results show that the algorithm has the advantages of good robustness,high accuracy,good real-time performance and so on.

关 键 词：汽车仪表 指针检测 最小二乘 图像细化 HOUGH变换 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]