Design And Implementation Of Multi Pointer Instrument Recognition Module

Dial indicator have been widely applied for producing and our daily lives due to its ability of anti-interference and facility in measurement.As functions mainly in measuring slight displacement,dial indicator will presents error with a high-frequency usage.Accurate calibration of dial indicator,considered as a significant procedure,is highly related with the quality and safety of products.Confined by the Accuracy test regulations which constituted by AQSIQ,an effective dial indicator should be calibrated within 12 months.The present accurate calibration mostly observed and recorded by naked eye and will drive eyes of observer tired out with intensive observation,and therefore,leads to erratic errors that impact on the accuracy.In this paper,the design of multi pointer instrument recognition module includes image acquisition system,image preprocessing system and instrument reading recognition system.The multi-pointer indicating meter recognition model(MPIMRM)uses the industrial camera,the optical lens,the illumination light source and so on to build the hardware platform of acquisition system,and the computer software is used to realize the image processing and reading recognition.Through the combination of artificial vision replacement and image automatic transformation,the image information is converted into digital information to get the meter reading.The identification module used in the automatic verification system of dial indicator,eliminate eye reading visual fatigue can also improve the reading efficiency and accuracy.The meter image is collected by CMOS industrial camera which connected with PC through the USB and transmitted to PC,and afterward,being processed by image pretreatment model."Hough transform"and"mage matching recognition algorithm"are two algorithms of digital identifying model in this paper and simulated through different software for comparison.Hough transform detects the line of pointer in the pretreatment image,calculates the slope and the angle of pointer,calculates the meter value through the certain ratio between the value and the angle of pointer moves."Pattern matching recognition algorithm" extracts area-of-interest from the meter as a model,matches pixel through Pattern matching sub-model of LabVIEW.Identify the angel of pointer moves with the position of zero scale,pointer and centre of the meter,and then calculates the final value.With experimentation,the result shows the pattern matching algorithm to meet the design requirements,the maximum recognition error is less than 0.003mm can be positive and negative,which is qualified for application.Friendly HMI coordinates with dial indicator calibration system to convenient users.