LiDAR Localization using reflective properties in harsh environment that interfere with LiDAR sensing

Abstract

LiDAR localization is an essential technology for autonomous robots. By positioning, the robot can determine its location and drive toward a mission point. This paper presents a novel localization method to overcome the limitations of lidar sensor data in harsh environments that interfere with lidar sensing, such as an environment in which haze is scattered or surrounded by glass structures. Unlike the existing method that uses the geometrical features of lidar sensor data, in this paper, we propose a feature that is characteristically expressed even in the harsh environment above by utilizing the reflective feature of lidar, and localization method using this feature is also proposed.