Research On Non-contact Detection Of Ball Screw Thread Parameters Based On Image Processing

Ball screw is currently widely used in high-end machine tools,in order to ensure the reliability of ball screw,therefore ball screw parameters need to be measured with high accuracy.Due to the low efficiency of manual testing of ball screw parameters,scholars now use non-contact image processing techniques to measure ball screw parameters.However,the CCD camera can not take the ball screw thread profile and therefore can not measure the parameters directly on the ball screw image.In this paper,ball screw geometry model was established and explained ball screw tooth projection occlusion,then solve ball screw profile equation and ball screw tooth profile curve equation,finally get ball screw tooth profile difference equation.The genetic algorithm was used to calculate the optimal value of ball screw tooth profile difference equation.The curve of ball screw tooth profile was fitted by the least squares method.By studying the ball screw parameter non-contact detection system,the lighting system and the imaging system were designed to ensure the high image quality of ball screw.The ball screw image preprocessing,the outline extraction of tooth profile and measuring ball screw parameters process were completed by using LabVIEW software.In the experimental part,the calibration coefficient conversion and reconstruction of the tooth profile boundary of ball screw image were carried out and ball screw tooth screw was finally obtained.The error of detection system was analyzed and the result of ball screw parameter measurement was statistically analyzed.The experimental results show that the proposed method can accurately measure the parameters of ball screw by reconstructing the tooth contour of ball ball screw image.