Research On Vision Measurement System For The Thickness Of Grinding Wheel Layer

Grinding plays an important role in modern manufacturing industry.The grinding layer of wheels will continually wear during the manufacturing process.The grinding wheel must be replaced when the wear reaches a certain extent.Considering the high price of superabrasive wheels,replacing early will affect economic benefits,while replacing late will increase the risk of production safety and processing quality.The study on the measurement of the grinding wheel layer benefits to the manufacturing industry.This thesis studied on vision measurement to the thickness of the grinding wheel layer.Designed an vision measurement system calculating the full circle thickness.The main work of this thesis is as following:1.The design of vision measurement system.Propose a system layout in which the directions of the camera and mainshaft are the same.The view of the camera contaions a part of grinding wheel layer.The wheel will rotate during the measurement.The thickness of the grinding wheel layer could be measured.2.The study of imaging process algorithm.The algorithm contains image preprocessing,binarization,horizontal grayscale projection,location of the ring width and subpixel subdivision.Calculate the pixel number of thickness from an image containing a part of grinding wheel layer.3.The study the super resolution of grinding wheel images.Build the super resolution model based on CNN.Prepared the training dataset.Involve transfer learning strategy.The super resolution model could cut down the experimental standard deviation about 10% compared with interpolation.Examine the effectiveness of the measurement system by various kinds of experiment.The experimental standard deviation is less than 15 um when collecting more than 120 images per circle,which satisfies the need of this project.