Research On Automatic Recognition System Of Pointer Meter's Reading

Pointer type instrument not only has the advantages of simple structure,but also has the advantages of waterproof,dustproof,anti electromagnetic interference etc.Therefore,in the field of petroleum industry,pointer pressure gauge has a very wide range of applications.At present,the pointer meter reading is mainly based on the human eye observation.In order to improve the rapidity,accuracy and efficiency of detection,this paper designs a system which can automatically identify the petroleum industry commonly used pointer pressure gauge.This system mainly uses the image processing technology based on machine vision to get the pressure value.The system uses a 5 million pixel CCD camera,firstly preprocess the collected instrument image,including the following three steps.First,the median filter is used to remove the salt and pepper noise in the image,and the edge of the image can be well preserved.Then the piecewise linear transformation is used to enhance the information of the important interest parts such as the pointer and the scale line.Finally,we use Canny operator to detect the edge of image.According to the shape and position of the dial,the edge of the dial is extracted.Then the contour of the dial with the pointer and the scale line is segmented by the edge.Through the preprocessing,the image of the dial is obtained,and then the pressure value can be determined by the position relation of the pointer and the scale line.The distribution of seismic pressure gauge calibration line is not uniform,and there are transparent oily liquid will block part of the scale line.Therefore,the traditional angle method or distance method is not suitable for this system.This paper proposes an image normalization algorithm by two-dimensional interpolation algorithm,limiting the initial and end angles of interpolation search.Then can remove dial text in the image and identify the circular dial image into rectangular image.Then the two pointer at different positions of the image to produce a pointer free dial and smooth the stitching.Then,a Zhang thinning algorithm is used to get the single pixel scale line template to facilitate the extraction of the scale line.Then through improved Hough Transform detect the scale line,and fill out the scale line which is obscured by damping liquid according to the position of the scale line.So that a complete calibration line template is obtained.At the same time,the coordinates of each scale are included in the array.Finally,the position of the pointer is obtained according to the difference method,then can get the coordinates of the pointer.According to the dividing value of the dial,the uniform scale reading can be obtained by using the linear relationship,by using the piecewise linear relation,we can get the gauge reading of the non-uniform scale instrument.Finally,this article uses Microsoft Visual Studio 2010 to design a dialog based MFC application.According to the pressure gauge calibration requirements,the program uses a modular design,including three modules: registration module,identification module and analysis record module.Through comparative analysis of the recognition of the pressure and the digital instrument generate pressure gauge calibration report.