Research On Industrial Pointer Pressure Instrument Detection

In the factory verification process of industrial pointer instruments,accurate automatic reading of the instrument can be achieved,which can increase the efficiency of instrument verification.Compared with the inefficient method of traditional artificial pressure and human eye reading detection method,the machine vision-based pointer pressure meter reading detection method has the advantages of fast speed and high accuracy.This article designs an industrial pointer instrument detection system for the inspection site according to the instrument factory inspection specifications and relevant national regulations.Firstly,a pointer pressure instrument detection experiment platform was designed.According to the national pressure instrument verification regulations and the verification method of the instrument manufacturer,a controllable pressure generator is used to provide a controllable standard pressure for the instrument verification.It is equipped with standard digital pressure gauges,industrial cameras,lighting sources and other equipment.The upper computer software of the experiment set up a pointer pressure meter reading detection platform.Then,the attitude correction algorithm of the instrument panel is studied.In view of the possible existence of dust during the detection process,median filtering is performed on the collected instrument images to improve the image quality.Aiming at the problem of attitude errors in the placement of industrial cameras and the installation of the instrument to be inspected,an improved image registration algorithm for grid movement statistics was used to extract the feature points of the standard instrument template and the instrument under test,and template matching was performed based on the extracted feature points.Calculate the projection matrix of the two,and correct the attitude of the dial according to the obtained projection matrix to eliminate the attitude error.Secondly,the dial division and dial circle positioning are studied.For possible background interference,the processed dial image is segmented by OTSU threshold to eliminate the interference of the background area,and the image of the dial part is extracted.Hough circle detection is performed on the extracted dial image to find the circle with the largest radius and locate the position of the dial circle.Finally,the method of extracting dial hands and reading the hands was studied.For the pointer extraction,various interference straight lines in the dial are used to expand the dial image after OTSU binarization to eliminate the interference part of the dial area and retain valid pointers to achieve the pointer extraction in the dial.For the straight line detection of the pointer,there is no need to traverse the entire image,the algorithm combines the area where the pointer exists under the current pressure value,and performs straight line detection on the image within a certain angle range.According to the design drawing of the instrument,an angle-pressure identification model was established.Under this model,the angle of the pointer straight line was calculated to obtain the pressure reading of the instrument.