Research On Reading Method Of High Precision Pointer Instrument And Design Of Verification System

The high precision pointer instrument is an indispensable measuring tool in industrial production due to its high precision,wide scope of application and low price.However,wear caused by operation and changing environment may affect the measuring accuracy of the instrument in the long term use of instrumentation.Therefore,our country stipulates that the pointer instrument must be alibrated regularly.Although automatic calibration of the instrument has become the research tendency of instrument verification industry,which has the capability to liberate people from complicated and exhausting work,these systems have not been applied widely at present,with low accuracy,poor adaptability,recurrent visual error and high prices.In order to solve these problems,this paper launches a series of research,and the main contents are as follows:First of all,a variety of image processing and feature extraction algorithms for pointer meter recognition were studied in depth.The image was processed step by step,and the scale line and the edge feature of the pointer were extracted.We proposed a method combining LSD(Line Segment Detector)and least squares,which can get the position of scale lines and the pointer more accurately and fit the straight line equation.Secondly,we studied the method of reading identification for the high precision pointer instrument.In order to solve the visual error problem,we analyzed the viewpoint error model in which there was an angle between the lens plane and the instrument plane as well as a displacement between the lens optical center and the pointer rotation center,and gave the calculation method of the viewpoint error.Futhermore,we proposed a reading method based on “three points in the same line”,using template matching technique to locate the coordinates and to obtain the intersection of the calibrated scale line and the scale arc.After calcul ation,the visual system is moved right above the calibrated scale line,making the visual system the center,the pointer,and the pointer's shadow in the mirror in the same line.Then we use the magnified distance method to get the reading result.This method improves the accuracy and versatility,with the capacity to calibrate instruments of different dimensions.Based on the above theoretical research,this study designed and implemented an automatic verification system for 0.5 stage high precision poin ter instrument.Completed the construction of the hardware platform including the standard source,the precise displacement platform and the visual system,we developed each software module and designed the interface of the system.It was found that light intensity has an effect on the results of the system,therefore,we designed an intelligent lighting system.The linear relation between the mean gray value and the exposure time is fitted by experiments and the average gray value of the best illumination environment is determined in advance.When the intensity of illumination changes,we can estimate the best exposure time of the camera and make the system adaptive to the light environment.Finally,a large number of experiments were carried out to verify the system.Experimental results show that the system has good accuracy,stability and versatility.Meanwhile,the verification efficiency is significantly higher than the manual verification efficiency.