A Method for Detecting Human by 2D-LiDAR

Abstract

Small robotic carts have been developed that can be used as transport, guide, delivery, or patrol robots. These carts move autonomously using a localization system, obstacle detection, and maps. We focus on a patrol system to detect suspicious but harmless individuals, such as elderly wanderers. We assume that the patrol carts have GNSS as their localization system, laser imaging detection and ranging (LiDAR) for obstacle and human detection, and a BLE device for authorized human detection. In this study, we propose a method for detecting humans using 2D point cloud data with LiDAR.