In order to achieve the detection of r ailway intrusion into foreign bodies,and overcome the shortcomings of low reliab ility of camera detection and single detection data of lidar.A foreign object de tection system combining camera and 2D radar was designed to visualize the software test results.Based on the two-dimensional lidar,the target recognition method based on condensed hierarc hical clustering algorithm is established.The target recognition experiment is carried out for a single perso n and three people.At the same time, the indoor and field single targets based on Kalman filtering are established.T he body and multi-target tracking method is used to perform target tracking experiments on multi-targets in the c ase of tracking loss,coincidence and occlusion.Finally,the two-dimensional lidar and the camera are combined with in formation to test the single target and multiple targets.The results show that the point cloud data collected by th e laser radar can realize the recognition of the target body through the condensed hierarchical clustering alg orithm and the denoising process. The Kalman filter method can be used to track the target body in a short time,w hich is lost,occluded or When other targets are combined,tracking again,the results of the fusion of 2D lidar and camera information can better obtain the position and boundary of different targets.The research results can provide theoretical reference and basis for the intelligentization of railway safety detection system.