Abstract:Precision bearings are widely used and high requirements are required for their accuracy. The defect of bearings surface necessary to detect the bearings defect. At present, the defect detection of the bearings mainly depends on people, however, when the defect is less than 0.075 mm, it is difficult for people to identify. The charge coupled device (CCD) camera and image processing techniques were adopted to design a detection method on line, which could greatly improve the detection efficiency and accuracy. Finally, the BP neural network was used to classify the defects.The experimental results show that the classification accuracy is up to 92.7%, which meets the requirements of the industry.