Safety Support Network System Using Overhead Sensing by a UAV Equipped with 3D LiDAR

Preventing accidents at intersections is a major issue in realizing safe and secure smart cities. Existing research has proposed systems for observing objects by utilizing UAVs equipped with cameras. However, a major challenge remains in recognizing objects under low-light conditions. Additionally, methods using three-dimensional LiDAR, which can observe the three-dimensional structure of the real world in detail have also been proposed. However, when LiDAR is mounted on the ground or in vehicles, occlusions (blind spots) may temporarily occur due to the movement of the vehicle or other nearby vehicles, which affects accuracy of the target. Therefore, we propose a system that uses a three-dimensional LiDAR mounted on a UAV to detect and identify vehicles and pedestrians passing through an intersection by sensing the observation area without blind spots from the sky. In the proposed system, the processing from data acquisition to analysis is executed on an edge device mounted on the UAV, allowing real-time feedback of the analysis results to the physical environment. In addition, by combining continuously collected three-dimensional point-cloud data from LiDAR based on inertial navigation data obtained from an Inertial Measurement Unit (IMU) and Global Positioning System (GPS) receiver, high-precision self-localization and 3D map generation are achieved.