Research On Precision Measurement Of Cable Thickness Based On Subpixel Image Processing

With the rapid development of production,science and technology,the demand for wires and cables in our country is also growing rapidly,which requires higher accuracy,speed and automation of measurement methods.The traditional cable thickness measurement method can hardly meet the new requirements.A complete non-contact image measurement software system is designed to meet the requirements of cable sheath thickness measurement system proposed in GB/T 2951-11-2008.The software system has sub-pixel accuracy.Four aspects of image preprocessing,pixel-level edge detection,sub-pixel edge detection and corner detection are studied in depth.Three sub-pixel relocation algorithms,Hueckel,graymoment and Zernike orthogonal moment,are deeply studied and implemented.An improved sub-pixel relocation algorithm of Zernike orthogonal moment is proposed,which improves the real-time performance and anti-noise performance of the algorithm.The error of the algorithm is within 0.015 mm for non-corner edge location.In view of the large error of traditional algorithm in corner position measurement,a sub-pixel relocation algorithm based on corner detection is proposed.The error of the algorithm in corner position is reduced from 0.03-0.06 mm to 0.015-0.025 mm.The specific contents of this paper are as follows:1)The hardware of the measurement system is deeply studied and introduced.The lighting system,cable sheath slice,CCD camera and image acquisition card are briefly analyzed.A fast and accurate method for calculating calibration coefficient is introduced.2)The process of image acquisition and transmission is thoroughly studied,and the noise sources of the image are introduced.Then several commonly used spatial filtering algorithms are introduced.Finally,a new adaptive weighted median filtering algorithm based on edge protection is proposed.The experimental comparison of these algorithms under various noise conditions is made with Lena image,and the validity of the new filtering algorithm is verified.3)Introduce the relevant algorithms and specific steps of pixel-level edge detection in detail,then study three classical sub-pixel relocation algorithms in depth,and compare the three algorithms through MATLAB simulation images,and analyze the shortcomings of the three algorithms.4)The requirements of cable sheath thickness measurement system proposed in GB/T 2951-11-2008 are introduced in detail.Then an improved Zernike orthogonal moment sub-pixel relocation algorithm is proposed.On the basis of pixel edge detection,Zernike orthogonal moment sub-pixel relocation algorithm is used,and 7*7 template is used to improve the noise resistance of the algorithm.Finally,the improved algorithm and several other algorithms are applied to the slice of cable sheath,and the thickness of the slice is measured.Six related thickness values are measured and compared with the actual thickness.The effectiveness of the improved algorithm is verified.5)In view of the large error of the general sub-pixel relocation algorithm in the corner position,a sub-pixel relocation algorithm based on CPDA corner detection is proposed.This algorithm is introduced in detail.Then it is applied to the slice of cable sheath.Thickness of the slice is measured,and six related thickness values are measured.The comparison with the actual thickness is made and verified.The effectiveness of corner sub-pixel relocation algorithm;...