Research On Depth Measurement Technology Of Thermal Protector Calibration Point Based On Machine Vision

Machine vision detection technology is favored by researchers because of it's strong robustness,good real-time performance and high measurement accuracy.In recent years,with the rising of intelligent manufacturing,the role of machine vision detection technology in industrial production has gradually emerged.As an important over temperature and over current protection device in electrical equipment,the position and depth of the calibration point of the thermal protector directly affect the quality of the product.However,the shell of thermal protector is made of metal and the calibration point is small,which puts forward higher requirements for the current detection technology.In order to realize the center position detection and the depth measurement of the calibration point of the thermal protector,this subject is based on the existing machine vision detection technology.The specific work is as follows:(1)An experimental platform for measuring the calibration points of thermal protectors is built.Firstly,based on the principle of three-dimensional measurement of structured light,the experimental platform of the thermal protector calibration point detection system is designed.Then according to the platform design and measurement indicators,the selection of hardware equipment is completed.Finally,the relevant theory of system calibration is studied,and the calibration experiment of the detection system is carried out,and the calibration results are analyzed.(2)Aiming at the problem that the background of calibration point of thermal protector is complex and its volume is small,its edge shape is irregular,and the traditional circle detection algorithm is not ideal for the center detection of calibration point,an improved random Hough transform center detection method is proposed.In order to improve the detection accuracy of irregular circle,this method replaces random sampling by area sampling and adds the best circle screening link of circle center 8 neighborhood.The experimental results show that compared with the traditional random Hough transform circle detection algorithm,the deviation coefficient of the improved random Hough transform center detection algorithm is reduced by 9%,and the detection speed is multiplied.(3)Aiming at the problem of reflecting light on the surface of thermal protector,a method of extracting the center line of structural light based on principal component analysis and region growth is proposed.Firstly,the mask method is used to process the image with pseudo light stripe,then the principal component analysis method is used to solve the normal direction of light stripe,and finally the region growing algorithm is used to extract the center line of structured light.Experimental results show that the proposed method can accurately extract the structural light center line of reflective metal surface.In terms of extraction speed,the algorithm in this paper is 5times as fast as Steger's method;in terms of extraction accuracy,compared with gray center of gravity method and Steger's method,it improves 1.081 pixel and 0.856 pixel respectively.(4)The 3D reconstruction of the thermal protector is completed.On the basis of the previous research,the three-dimensional point cloud data of the surface of the thermal protector is obtained.Through the filtering,simplification and triangulation of the three-dimensional point cloud data,the three-dimensional reconstruction of the thermal protector is completed.According to the three-dimensional model,the depth data of the calibration point is obtained,and the error of the measurement result is analyzed.The measurement error of the whole detection system is within 0.05 mm,which reaches The measurement requirements of manufacturers have certain practical application value.