Radon transform-based indoor localization control for quadrotor

摘要

Quadrotor aircrafts have been widely used in many fields, but its methods of indoor localization control based on visual recognition are relatively few. For identifying and localizing a target with high speed and accuracy, visual signal is employed as feedback to control quadrotor. Considering Radon transform can extract the shape feature of the target and decrease the computational complexity efficiently, this paper proposes a novel method SFRT (Shape-Finite Radon Transform) with visual feedback for image processing. A series of algorithm optimization tests have been completed for the fight characteristics of quadrotor. Then the feasibility of SFRT is verified with the recognition accuracy which can reach more than 90% in setting conditions indoors. (C) 2016 Elsevier GmbH. All rights reserved.