研究了一种基于机器视觉的汽车仪表盘自动检测系统.该系统通过机器视觉的方法,采用OpenCV编程环境,采集有效的汽车指针仪表盘图像,并对其进行处理,识别出仪表刻度值与指针的位置和重合程度,从而判定仪表指示准确性是否符合要求.系统包含了光源的配置、摄像头的选取和其横纵竖三向的移动、图像的预处理以及图像的识别判定等部分.测试证明,该系统能有效地自动读出指针仪表的示数并判断其是否准确,识别速度和精确度与国外类似产品相比得到了提升.%A system of automobile instrument inspection is designed, which is based on the OpenCV programming environment and uses machine vision technology. At first, the system collects effective images of the automobile instrument, and then the system will do some necessary image processing and recognition in order to readout the value of the instrument calibrations. Through those the system could judge the accuracy of the instrument whether it meets the requirements or not. The design contains several parts such as configuration of light source, selection and horizontal longitudinal and vertical movement of camera, image preprocessing and image recognition and so on. Experiments show that, this system can effectively read out the value of the vehicle pointer instrument and points out the dial precision by an automatic way, comparing with the similar products abroad the distinguishing speed and accuracy have obvious improvement.
展开▼
