Abstract: In order to solve the problems of hole edge quality detection and burr removal automation in the processing of sheet metal parts, the method of rapid detection of punching quality of sheet metal parts by using machine vision technology and the solution of automatic grinding of burr edge by vision-guided electric grinding head are studied. Firstly, the digital images of sheet metal holes with machining defects are collected, and the image processing detection algorithm is designed by using OpenCV vision software library to realize the center point positioning and edge burr detection of sheet metal holes. Secondly, the vision guided automatic grinding method for the burr edge of the punching hole is studied, the vision calibration and network communication test of the manipulator are carried out, and the guidance of the manipulator and the control of the automatic grinding motion path are realized. Then, the PC control software designed by C# is used to integrate the functions of image processing, communication control and visual guidance automatic grinding. Finally, with the six-axis industrial manipulator as the core, the visual inspection device and automatic grinding device are carried out, and the system is completed. After the test, the burr detection rate and grinding qualified rate of the system reached 85.7% and 100%, respectively, and the system has preliminary capacity for industrial production, and the system has the characteristics of light weight, intelligence and high integration, which can be widely used in production scenarios such as processing quality detection and automatic edge grinding of various sheet metal parts.