Measurement method of screw thread geometric error based on machine vision

Abstract

In order to solve manual operation and low detection efficiency in the measurement of thread geometric error, machine vision and optical enlargement are adopted to measure the high-precision geometric error of thread. For screw thread edge image characteristics, a new method using 67.5° and 112.5° improved Sobel templates to obtain the edge of image was put forward, and we calibrated the system using self-calibration method. The geometric error measurement system of screw thread was designed. The geometric error of screw thread, such as pitch, angle and diameter of screw, were calculated by extracting the coordinates of image edge. The system can be concluded that the method of machine vision, whose linear precision is less than 10 μm, can be used to detect the comprehensive parameters of screw thread. The designed measurement system was intuitive, efficient and easy to operate. It can be used in modern manufacturing and has values in generalized application.