Detecting and Localizing Objects on an Unmanned Aerial System (UAS) Integrated with a Mobile Device

Abstract

There is a strong need for Unmanned Aerial Systems (UAS) that can navigate autonomously using visual information for GPS denied and other degraded environments. In this paper, a system architecture is proposed that utilizes visual-inertial odometry (VIO) and object recognition and localization for a quadrotor UAS. It also incorporates communication with an Android Team Awareness Kit (ATAK) device, allowing users to share situational awareness of the UAS in real-time. The system architecture is validated through experiments in which a human is detected and localized as the object of interest.