Research And Implementation Of Key Technology For High Precision Automatic Detecting Device For Tool Flaw

When the power or mobile phone battery pole piece is cut,the micro-level flaw on the tool will cause the pole piece burr,which is a direct danger source for the battery fire.However,the quality inspection of tools at home and abroad is usually observed by the human eye with a microscope eyepiece.The process is time consuming,the intensity of the eye is extremely large,and the detection accuracy and repeatability are poor.In order to solve this problem,the paper designed an automatic inspection system for tool flaw.The detection system is mainly composed of image acquisition,image stitching,flaw detection,data management,etc.The machine vision related technology is used to detect the tool flaw generated when it is shipped or used.In this paper,the research on the tool defect detection and microscopic image stitching algorithm is carried out.The main research contents are as follows:(1)In regards of the problem of low search speed of contour matching in image stitching process,a regional frog leaping search algorithm is proposed.The optimal value for matching contour similarity is quickly achieved,through the four stages of the initial distribution of frog group,the update of tweet decibel,the jump of the frog and the update of the decision domain.The experimental results show that the regional frog leaping search algorithm has better resolution and convergence speed.(2)For issue of rare features and difficult stitching in microscopic images,a stitching algorithm based on the similarity between image contours is proposed.Two kinds of indicators are introduced to measure the similarity distance of the contour curve,and the similarity and the discrete distance of the image curve are calculated respectively,which replaces the traditional SIFT or SURF feature matching method to improve the matching precision in image stitching.The experimental results show that the MSED distance as an image matching index has higher stitching precision and stronger robustness.(3)For issue of low accuracy and poor repeatability of the flaw detection in the microscopic image of the tool,a boundary-sensitive tool flaw detection algorithm is proposed.The regular constraint term is introduced in the objective function of the least squares fitting method,which improves the fitting ability of the reference fitting line to the tool repair profile.The elastic adjustment space and the boundary sensitive network are constructed to adjust the boundary position of the flaw target.The boundary-sensitive flaw detection model is established by the above improvement,and the results show that the algorithm has higher detection accuracy and lower miss detection rate in tool edge flaw detection.The comparison experiments show that the proposed method of flaw detection has a good recognition effect in tool flaw identification,and the practical application of domestic leading enterprise B shows that the proposed algorithm not only has high detection accuracy and strong robustness.At the same time,it meets the requirements of online real-time detection...