Aiming at the problem that the automatic inspection system of conveyor running state has poor accuracy in extracting the edge of conveyor belt, and it is difficult to ensure the safe operation of the automatic inspection system, a safety identification technology of conveyor belt edge based on laser detection is proposed. Taking lidar as the core, the point cloud preprocessing technology, point cloud plane segmentation technology and conveyor belt edge recognition technology of monitoring image are mainly studied to ensure the accuracy of image recognition. At the same time, the application reliability of this technology is studied by means of simulation analysis. The results show that when the shielding rate of the underground environment to the lidar is less than 35 %, the extraction accuracy of the system to the edge of the conveyor belt reaches 78.29 %, which is 24.79 % higher than the current recognition technology. It plays an important role in ensuring the inspection accuracy and safety of the automatic inspection system for the running state of the conveyor.