Analysis of Abnormal Transmission Synchronous Belt of Industrial Robot Based on Visual Inspection

In view of the phenomenon that the synchronous belt of industrial robot will jump teeth and slip in use, an image acquisition system based on MATLAB is designed. The image collected by the macro camera is preprocessed to obtain the general contour boundary. In order to obtain the high accuracy synchronous belt tooth profile data, it is necessary to fit the data. Harris corner detection screen boundary feature points, through cubic spline interpolation, digital filtering algorithm to get the best contour curve. Comparing the tooth profiles of different synchronous belts can directly reflect the cause of abnormal transmission of synchronous belts. Compared with the traditional manual detection, avoid the secondary measurement error caused by contact. The tooth shape of the incoming synchronous belt is controlled by visual inspection, which is easy to operate, the measurement accuracy is controlled at the pixel level, and the data is reliable.