Visible Light Communication on LED-equipped Drone and Object-Detecting Camera for Post-Disaster Monitoring
Hiroki Takano, Daisuke Hisano, Mutsuki Nakahara, Kosuke Suzuoki, Kazuki Maruta, Yukito Onodera, Ryo Yaegashi, Yu Nakayama
Abstract
This paper proposes the concept of a visible light communication (VLC) system with LED-mounted drones for post-disaster monitoring and reports the results of the experimental feasibility evaluation. Post-disaster monitoring is a critical social issue given that in the case of a large-scale disaster, communication failure can occur, and the power supply can be interrupted. A VLC system is useful for investigating the disaster situation, can assist in life-saving activity by lighting the area, and allows for communication from the disaster area to a base station. In this study, the VLC system consists of LED-mounted drones and an image-sensor-based receiver. The drones are videoed and subsequently detected by the camera with a YOLOv3-based convolutional neural network (CNN). After detecting the position of the drones, the light signal is demodulated. This paper reports the following feasibility evaluation results; 90% accuracy in drone recognition and a low bit error rate with the distance of up to 80 m between the drones and the camera.