Human Intrusion Detection System using mm Wave Radar

Due to the giant advances in 5G technology and the Internet of Things, target detection and identification play increasingly significant roles in daily life. In this paper, a human intrusion detection system with high robustness based on mm Wave radar is built to solve the problem of regional rapid automatic human intrusion detection in complex environments. In the system, we propose a novel approach to rapidly identify human intrusion targets, mainly including feature extraction using 2D-FFT, MTI, CFAR, and clustering, and target identification based on SVM and point cloud. Experimental results indicate that the system has a success rate of exceeding 85% for human intrusion detection under different complex conditions, and has strong robustness in most scenarios.