On-machine Measuring The Profile Accuracy Of Profile Grinding Wheel Based On Machine Vision

Profile grinding is widely used in the profile machining of various precision and complicated curve components.It is necessary to maintain the grinding wheel when it wears to a certain degree.At present,there is no efficient and accurate way to in-situ measure the profile of grinding wheel.Therefore a method for on-machine measuring the profile accuracy of grinding wheel based on machine vision is studied.This method focuses on in-situ detection of the wheel wear state as well as measuring the wheel dressing accuracy,in order to optimize the dressing opportunity and ensure the dressing accuracy.The main contents are as follows:(1)A machine vision based grinding wheel profile measurement method was proposed to achieve fast,accurate and in situ detection of the profile accuracy.Experiment platform for measuring the profile accuracy of grinding wheel is designed and constructed to study the feasibility of on-machine detection of the profile grinding wheel's profile accuracy.(2)CCD camera calibration technique is studied.On the basis of comparing different calibration method,four-point center calibration method is used to do the image calibration experiment,which has high accuracy and maneuverability.Finally,different standards distance templates are used to test the calibration accuracy of the four-point center calibration method.(3)Edge detection of the grinding wheel's profile is studied.Theoretical analysis and experimental study are done for the image preprocessing noise reduction.The results show that the median filtering can suppress noise effectively and preserve image detail at the same time.A method is proposed that uses canny edge detection as a rough detection and Zernike moments based subpixel edge detection as fine detection.Experimental studies show that this edge detection method has stability characteristic which is insensitive to noise.The edges obtained are smooth,which improves the edge detection accuracy.(4)Profile feature of the grinding wheel is abstracted.The least squares circle fitting method is used to extract the features of the grinding wheel's profile.The radius error and roundness error are analyzed by comparing the abstracted feature with the ideal profile feature.The profile of grinding wheel was measured by CMM to acquire the experimental profile data.The measurement data were analyzed and the deviation between vision measurement based result and CMM based result is calculated.The result showed machine vision based detection of the wheel profile can well meet the required measurement accuracy.