Research On Grinding Wheel Profile Measurement Based On Machine Vision

In the grinding process,the profile of grinding wheel directly affects the shape accuracy of workpiece.Therefore,the accurate profile measurement and parameter evaluation of grinding wheels of different shapes are of great significance for the evaluation of grinding wheel machining accuracy,the compensation of machining error and the guidance of grinding wheel dressing.According to the requirements of grinding wheel profile measurement and parameter evaluation,a grinding wheel profile measurement method based on machine vision is proposed in this paper.On the basis of research on autofocus algorithm and image mosaic algorithm,a grinding wheel profile measurement system is established.Finally,non-contact and efficient profile measurement of grinding wheels of different shapes and sizes is realized,and the profile parameters that affect the machining accuracy are evaluated.1.According to the requirements of grinding wheel profile measurement,a measurement method based on machine vision is proposed and a corresponding profile measurement system is designed.The system error is modeled and analyzed.Based on the error model,an actual measurement system is built and the system error is effectively suppressed and compensated.2.The ROI-based Brenner sharpness evaluation function is proposed to obtain the sharpness evaluation curve of grinding wheel profile image sequence,and the experiment shows that this function has higher efficiency,better unimodal and stability than other traditional functions.Combining the sharpness evaluation function with the developed adaptive step search algorithm and the improved Gaussian curve fitting,the fast and accurate search for sharpness evaluation curve peak is realized and the best focusing position is obtained,which achieve autofocus of the grinding wheel profile.3.By analyzing the noise type and intensity of grinding wheel profile images,the best denoising method is selected to preprocess the image to get high-quality images.A profile template matching based on the profile search algorithm is proposed to realize the image mosaic of grinding wheel cross-sectional profile.The experiment shows that this method has higher accuracy and efficiency than other commonly used image mosaic algorithms,and is more suitable for grinding wheel profile image mosaic in this study to obtain complete grinding wheel profile information.4.The two-dimensional and three-dimensional profile measurement and parameter evaluation of cylindrical and hemispherical grinding wheel are realized by using the measurement method based on machine vision.The feasibility and accuracy of this method are proved by comparative experiments.