Research On Milling Cutter Beating Tensile Detection Algorithm Based On Machine Vision

Machine tools are the mother of industry and milling cutters are an essential component in the machining of CNC machine tools.The accuracy of current CNC machine tools has reached a high level,but the accuracy of the parts required for various precision instruments is now becoming higher and higher.The runout of milling cutters can cause a reduction in the machining accuracy of the workpiece,while the wear and tear of the tool can also increase,ultimately leading to higher machining costs.Not much research has been done on milling cutter skip detection.This paper takes milling cutters as the research object,and uses machine vision to perform research on the technology related to milling cutter runout detection,so as to achieve the non-contact measurement of milling cutter runout detection.According to different photographic scenarios,the corresponding algorithms are designed,and better results are achieved in milling cutter runout detection.The main research content of the subject is as follows.(1)A machine vision image acquisition system designed for the experimental conditions and the characteristics of the milling cutter.Suitable cameras and lenses need to be selected and key parameters such as the field of view,focal length and photo distance need to be calculated.To investigate the influence of the illumination system on the imaging quality.This paper conducts experiments from three perspectives:light angle,colour and intensity and in combination with image processing algorithms.Based on the experimental results,an illumination system suitable for the subject is designed.Due to the limitation of the shooting equipment,the camera used in the experiment will cause certain distortion during the imaging process.In order to correct the distorted image and improve the detection accuracy,the calibration and correction is carried out using the Zhang Zhengyou calibration method.(2)To address the problem of motion trailing that can develop when the camera is capturing a rotating milling cutter.An image quality assessment system is built and the SSIM and GMSD algorithms are selected to discriminate the quality of the images in order to improve the efficiency of detection and prevent false detection problems.Experiments are conducted on the SSIM and GMSD algorithms combined with the edge localisation algorithm.The experimental results show that the images evaluated by image quality can effectively detect the edge of the milling cutter and achieve accurate positioning of the edge of the milling cutter.(3)The most important aspect of milling cutter skip detection is the ability to localise the edge of the milling cutter.To address the problem of inaccurate edge positioning,a sub-pixel edge positioning algorithm with Zernike matrix and improved polynomial interpolation is used.The milling cutter edges are accurately localised and the corresponding experiments are carried out.The experiments show that the sub-pixel positioning algorithm can accurately check the coordinates of the milling cutter edge and can effectively achieve the detection of milling cutter runout.(4)In this project,the jump of a triple-edged milling cutter and a ball-ended milling cutter is to be checked.The detection of runout is carried out in three photographic scenarios:camera front view of the three-edged cutter,camera side view of the three-edged cutter and camera side view of the ball-ended cutter.The edge characteristics of the milling cutter in the images differ depending on the perspective of the shot.For this reason,the corresponding algorithms were designed as an experimental theory for detecting runout errors in milling cutters from both camera front view milling cutters and camera side view milling cutters.The experimental results show that the method used in the subject is effective and can detect the milling cutter runout accuracy better.