동적 객체 추적 및 제거 기능을 갖춘 라이다 기반 실내 지도 작성

Abstract

Dynamic objects in point cloud maps generated using light detection and ranging (LiDAR) sensors cause ghost trail effects and decrease the quality of the maps, affecting localization accuracy. This study proposes a precise three-dimensional mapping method using LiDAR sensors in the presence of dynamic objects in indoor environments. To achieve this, a LiDAR SLAM algorithm that builds a point cloud map using a vehicle is integrated with a tracking filter algorithm that analyzes the motion of targets surrounding the vehicle. Experimental results in an indoor environment using a mobile robot are shown to validate the effectiveness of the proposed method.