High voltage cables are widely used in power transmission.Testing the key dimensions of high voltage cables to help ensure their quality is an indispensable part of the high voltage cable preparation process.At present,the domestic size inspection of high voltage cables is still in the manual inspection stage,and the measurement accuracy and efficiency are low.This paper proposes and studies a method for measuring the critical dimensions of high voltage cables based on machine vision to improve measurement accuracy and efficiency.The main research work and results are as follows:1.Design and establish a subregional imaging device based on machine vision and mechanical motion scanning,and calibrate the imaging system to meet the high precision measurement requirements for key sizes of large size cables.2.The SURF algorithm with rotation invariance and scale invariance is used to achieve registration and high-precision stitching of subregional images,and an algorithm based on guided filtering for image processing is established to effectively preserve the edge features of the image while filtering out noise.3.Based on the single pixel edge of the cable shielding layer outline obtained by the Sobel algorithm,a least-squares fitting method based on roughness evaluation is proposed to obtain the diameter parameters of the rough cable contour.Further based on the Sobel-Zernike sub-pixel extraction algorithm,the contour sub-pixel promotion is realized,the sub-pixel coordinates of the shielding layer contour are obtained,and the method of searching and matching the shortest distance is used to calculate the thickness parameters between the contours.4.On the basis of the above devices,algorithm and software,a key dimensions measuring system for high voltage cables based on machine vision is established,and experimental tests are performed to verify its good characteristics. |