Automatic Error Detection And Adjustment Of Auto-Meter Based On Image Processing

The traditional Auto-Meter deteciont is carried on through a manual way,including the instrument reading,LED and LCD birghtness test.However,the precision can not be guaranteed because of great labor time.Therefore,automatic detection and adjustment become a very important research domain for the production of Auto-Meter.Automatic detection based on image processing can avoid all these problems.This thesis design a system for automatic detection and adjustment of Auto-Meter based on image processing.Firstly,the system is designed according to performance parameteres,including hardware and software design of R33 Cluster,solution of AS2 work station.Secondly,image processing algorithm is designed for AS2 work station,which is composed of two parts.One part consist of recognition of pointer reading,computation of deviation and adjustment angle.The other part consist of LED,LCD brightness and colour test.The image processing algorithm for the automatic detection of Auto-Meter includes pretreatment,edge detection,image division,pointer extraction,and deviation computation. Because time is a very important factor in this system,it is considered at every steps of algorithm,besides the quality of image processing.Furthermore,the program of hough transform algorthim is optimized to improve test precision and reduce time.Finally,the algorithm is tested under a site test condition.Improvement direction are given for the potential problems.Meanwhile,a prospect for automatic detection of Auto-Meter is given.