A Robust Real-Time Embedded Vision System on an Unmanned Rotorcraft for Ground Target Following

Abstract

In this paper, we present the systematic design and implementation of a robust real-time embedded vision system for an unmanned rotorcraft for ground target following. The hardware construction of the vision system is presented, and an onboard software system is developed based on a multithread technique capable of coordinating multiple tasks. To realize the autonomous ground target following, a sophisticated feature-based vision algorithm is proposed by using an onboard color camera and navigation sensors. The vision feedback is integrated with the flight control system to guide the unmanned rotorcraft to follow a ground target in flight. The overall vision system has been tested in actual flight missions, and the results obtained show that the overall system is very robust and efficient.