Research On Key Technologies Of Machine Vision Rapid Detection System For Disc Cutter Gap

The electrode burr makes the battery have a major safety hazard.Using a disc knife with a gap to cut the battery pole is one of the main reasons for the burr on the battery pole.The existing disk knife edge detection system has poor effect on removing the attachment on the disk edge,and the edge detection accuracy is not high and the efficiency is low,which is difficult to meet the production needs of enterprises.In order to solve these problems,this thesis studies the key technology of rapid detection system of gap based on machine vision.The main contributions of the present thesis are as follows:(1)An Attachment Removal Algorithm for Disk Blade Based on Concave Point Matching.Aiming at the problem that the attachment of the disc blade often leads to the failure of the disc knife edge detection,false detection data or detection failure,this thesis proposes a method to remove the attachment of the disc blade.Firstly,the edge image of disk blade is preprocessed by OTSU method and morphological method.Then,the convex hull algorithm was used to calculate the adhesion area of the attachment,and the sliding window calculation function was used to locate and match the concave point of the adhesion area as the best segmentation point.Finally,the optimal segmentation line between the attachment and the disc edge was calculated based on B-spline interpolation inference,so as to realize the attachment recognition and removal.Experimental results show that compared with the attachment removal model based on RESCAN,the proposed method improves the IOU,SSIM,PSNR and MPA by 7.87%,0.0227,2.1051 d B and0.1101,respectively.(2)Disc knife edge detection model based on Zernike-Hermite.Aiming at the problem of edge detection of disk edge,this thesis proposes a sub-pixel edge detection method based on improved Zernike moment.The method mainly introduces Sobel operator for rough edge detection,which reduces the computational complexity of Zernike moment sub-pixel edge detection,and then calculates the integrated gradient value of disk edge image.It replaces the identification parameters of the original edge extraction model.Aiming at the problem of rapid detection of disk knife gap,this thesis proposes a gap detection method based on Hermite interpolation.Firstly,shape parameters are introduced into Hermite interpolation to infer envelope fitting curve clusters.Then,the PSO algorithm was used to calculate and confirm the optimal shape parameters,and the optimal initial edge of the disk knife in the curve cluster was located,which effectively avoided the Undershoot problem of envelope fitting.Finally,the Riemannian integral difference between the edges was calculated to locate the gap area and calculate the relevant indicators.The experimental results show that,on the one hand,the smoothness of the disk edge extracted by the proposed model is better,which is closer to the accuracy of sub-pixel edge extraction,and the average time consumption is much lower than the classic Zernike moment edge extraction algorithm.On the other hand,the model in this thesis effectively avoids the Undershoot problem of the initial edge of pre-inference.Compared with the comparison experiment,the model in this thesis performs better in terms of detection accuracy,and the error rate can be controlled within 1.9%.In addition,the detection speed of the proposed model is faster,which can reduce the detection time by about 40% compared with the PRe Net model.