Machine Vision Based Study Of Wear Detection Method For Disc Milling Cutter

The cutting tool play an important role in mechanical processing,which directly affects machining quality,efficiency and the processing cost in production.In the process of CNC milling,disc milling cutter with combined blade is widely used more and more.Compared with the traditional one-piece tool,it is a mainstream in practical production because of its lower cost and accessibility.In production,the blade is changed regularly.However,many blades can still be used after replacement with the increasing usage.Previously,only manual experience is used to judge the state of tool wear,in which an effective judgment standard and quantitative results are not easily obtained.As a result,many tools have been eliminated without reaching their due service life,resulting in a lot of waste.The application of machine vision technology to blade wear state detection can effectively solve the problems of low detection efficiency,poor accuracy,inconsistent detection standards and high detection cost.This thesis takes disc milling blade as the research object.The method of machine vision is used to detect its wear degree.The details are as follows:Firstly,the causes and main manifestations of tool wear failure are analyzed in detail.Combined with the actual production of enterprises and according to the wear characteristics of the tool,a corresponding detection standard is formulated.Secondly,the hardware system used in tool testing is selected and designed,including camera,combined light source and experimental bench.The Zhang Zhenyou method for camera calibration experiment is carried out.Through the comparative analysis of light source and the combination experiment with various light sources,a suitable combination light source system is determined.The experimental bench is also designed.Thirdly,a series of image processing is carried out on the collected image,including gray level processing by gamma transform,noise reduction with Gaussian filter,image segmentation by OTSU threshold method,image morphology processing and minimum outer rectangle.Finally,the detection area of the image is framed.The image pixel size is also calibrated to statistically calculate the wear area of the milling cutter blade.The connected domain is processed for the damaged part of the milling cutter blade.The Canny operator is used to extract the damaged edge,and the damaged area and depth are calculated.MATLAB GUI is used to design the interface of the testing software to facilitate the presentation of the testing results.A further verification experiment is carried out,in which the 19 JPC universal tool mic roscope is used to measure practical the wear area of a milling cutter blade.Then the experi mental results are analyzed and compared with the presented machine vision detection resul ts.It shows the efficient and accuracy of the method.