Research On Bit Detection Of Grinding Wheel Shape And Position Tolerance Based On Machine Vision

The traditional grinding wheel testing process is as follows: workers disassemble the grinding wheel,transfer it to the testing platform for testing,reinstall the grinding wheel,and continue to work.Disassembly is likely to cause abrasion to the grinding wheel,and new errors will be generated during reinstallation.Meanwhile,disassembly and installation will also increase production and testing costs and time consumption.In view of the above problems,a method of grinding wheel in position detection based on machine vision is proposed.OPC interface technology is used to control the machine tool in position to adjust the grinding wheel alignment to solve the wear and error caused by disassembly and installation operation.Machine vision is adopted to replace human eye judgment and detection time to solve the judgment error caused by human eye fatigue.The grinding wheel image is processed to obtain the form and position tolerance detection results.The advantages above break the limitations of traditional grinding wheel detection,improve the detection accuracy and simplify the detection process,and finally achieve the in-place detection of grinding wheel shape and position tolerance.The main research contents are as follows:1.Research on grinding wheel alignment method and geometric tolerance algorithm.By analyzing the structural characteristics and testing requirements of the grinding wheel to be tested,the mathematical model of grinding wheel alignment is established,and the calculation method of adjusting the alignment offset of the grinding wheel is formed,which provides the source of parameters for the position adjustment of the grinding wheel by controlling the machine tool.The mathematical model of grinding wheel profile tolerance is established,and the calculation method of symmetry,cutting width and wear height of grinding wheel machining surface is formed,which lays a theoretical foundation for testing the profile tolerance of grinding wheel.2.Research on the key algorithm of grinding wheel detection.Image filtering,sharpening and binarization algorithms were compared to select the best grinding wheel preprocessing algorithm.Compared pixel level and sub-pixel level edge detection,Canny pixel-level edge detection was used for the initial positioning of image edge,and zernik-moment based sub-pixel edge detection was used for precise edge positioning of image.The combination of the two can achieve higher precision edge detection.3.The experimental platform was built.According to the characteristics of the grinding wheel to be tested and the detection accuracy requirements,select the camera and light source whose performance parameters meet the conditions,and design the appropriate lighting system;According to the testing process of grinding wheel,the human-machine interface is designed to monitor the state of grinding wheel and test results in real time.Complete the integration of software and hardware,and realize the integration of grinding wheel control,image acquisition and image processing.4.Camera calibration technology.Based on the camera imaging model and the conversion principle from world coordinates to pixel coordinates,the internal and external parameters of the camera are calculated,and the calibration parameters are used to correct the image distortion.5.Experiment and data analysis.The experimental results verify the feasibility of the system design.The measurement accuracy meets the accuracy requirements of the system.The grinding wheel form and position tolerance detection function based on machine vision is realized.The error sources are analyzed and the corresponding improvement methods are put forward to pave the way for system optimization.