Vision-Based Autonomous Landing for Small-Scale Unmanned Rotorcraft

Abstract

Autonomous landing is a challenging issue for unmanned aerial vehicles (UAVs). This paper presents a system that provides vision-based autonomous landing for small-scale unmanned rotorcraft. This approach utilizes dual image sensors facing downward and performs digital image processing techniques such as the blocked histogram matching and the template matching techniques to ensure the timeliness in altitude estimations. The experimental results show the applicability of these approaches in real-world flights.