The Pointer Type Instrument Indicates The Value Reading In A Complex Environment

Analog display instruments have irreplaceable advantages over digital instruments.Nowadays,although electronic technology is very developed,analog display instruments are still widely used in industrial production.With the rapid development of Industry 4.0,the traditional manual reading method has gradually been replaced by the visual reading system,and the application of machine vision in Analog display instruments usually requires a stable light source to ensure that the system acquisition environment does not have large fluctuations,which is often difficult in some old devices and inconveniently retrofitted equipment,and the establishment of a relatively closed system also brings a lot of inconvenience to personnel inspections,equipment inspections and replacements.Therefore the reading of analog display instruments’ value under normal lighting conditions is very practical.Under the conventional lighting conditions,due to the change of the light source environment and some shadows caused by the people walking on the meter dial,the traditional reading vision system is prone to errors in pointer extraction,and the existing research does not give a way to solve the problem when extraction errors.Compared to the use of HOUGH transformation positioning instrument panel,and then use edge detection,image segmentation,differential imaging and other methods to extract the connected component of the pointer,and then use the image refinement,HOUGH straight line detection and least squares fitting algorithm to extract the pointer.In this paper uses the manual setting method,by setting the position and size of the circle of the fitting dial,the position of the starting point and the ending point of the instrument reading value is determined on the arc,and finally through the statistical arc point and The gauge pointers are extracted from the features of the grayscale sums between the centers of the circles,and a pointer extraction error discrimination method based on threshold segmentation is presented.The algorithm implementation language of this paper adopts C++,the image processing library uses the open source OpenCV,the UI interface of the system is based on the Qt,it is convenient to manually set the position,size,and instrument range of the fit dial circle.the algorithm of circle and line is based on Bresenham algorithm,the pointer extraction which based on the straight line minimum gray value principle also based on the Bresenham algorithm.For the evaluation of the instrument’s indication,this paper uses the angle method and the proportion method to compare,and through the comparison between the obtained indication value and the standard value,the angle method is used to obtain the indication value,and the principle of error was analyzed for the indications obtained by the two methods.The indication of the error makes a principle analysis of the error.The images used in this article are images collected without the illumination of a professional light source.At the same time,experiments were conducted under the conditions of no light interference during the day,interference with light sources,and use of ordinary LED spot light as the light source under dark light conditions.The experimental results also proved the reliability and high anti-jamming capability of this algorithm.This paper provides a theoretical basis for the convenience of the smart meter reading system and the anti-interference ability of the system,and provides powerful help for the old industrial equipment to go to Industry 4.0.