Aiming at the problems of inconvenient measurement, low efficiency and weak human-computer interaction in traditional gear parameter detection, a gear parameter detection and evaluation system based on machine vision is designed by using virtual instrument technology, image processing technology and machine vision technology with LabVIEW as the development platform. The system is designed to collect gear images, image processing, system calibration, shaft hole center detection and gear parameter detection and evaluation. It can measure and evaluate gear addendum circle, tooth root circle, single pitch deviation fpt , total pitch deviation Fp and common normal length deviation. The 0-level standard gauge block is used to calibrate the system, and the designed system is tested and compared. The test results are compared with the design theoretical values of the gear to evaluate the accuracy of the system. The results show that the detection and evaluation system has a friendly interface, easy expansion, high efficiency and convenient operation, and can automatically save and playback the collected data and visually evaluate the gear parameters.