AUTONOMOUS SURVEILLANCE DRONE

Dhavalshree Khedkar; Akshata Kasar; Vishwas Shelke

doi:10.17605/OSF.IO/H2TBD

Authors

Dhavalshree Khedkar Information Technology Department, Datta Meghe College of Engineering, Airoli, Navi Mumbai, Maharashtra 400708, India
Akshata Kasar Information Technology Department, Datta Meghe College of Engineering, Airoli, Navi Mumbai, Maharashtra 400708, India
Vishwas Shelke Information Technology Department, Datta Meghe College of Engineering, Airoli, Navi Mumbai, Maharashtra 400708, India

DOI:

https://doi.org/10.17605/OSF.IO/H2TBD

Keywords:

Drone, Nvidia Jetson Nano,Microsoft Azure IoT hub, NVIDIA DeepStream SDK, YOLO.

Abstract

We present Autonomous Surveillance Drone, a new approach to serve purpose of UAV surveillance. Instead, just sending video stream from drone to display device at user's side we have developed complete system to get insights from video stream. Our system architecture is extremely simple and biggest reason for this is platforms provided by Nvidia and Microsoft, throughout this paper we are demonstrating all results/outputs using services and hardware provided by Nvidia and Microsoft. We are using the Nvidia's Jetson Development board to compile our inference part(object detection using YOLO-V3) along with stream and metadata fetch at remote terminal, Jetson nano is our IoT edge device, to connect edge device and the remote terminal we are using Microsoft Azure IoT hub (Cloud Service).