Zsedrovits, Tamás and Bauer, Péter and Zarándy, Ákos and Vanek, Bálint and Bokor, József and Roska, Tamás (2014) Towards real-time visual and IMU data fusion. In: AIAA Guidance, Navigation and Control Conference. AIAA, Washington, x-xx. ISBN 978-160086962-4 10.2514/6.2014-0604

Text Zsedrovits_1_2704366_z.pdf Restricted to Registered users only Download (1MB) | Request a copy

Abstract

In this paper sensor models, implementation aspects, and simulations and measurements obtained during the development of a fixed wing mini UAV are presented. The aim is to develop a low cost UAV which can be integrated into the national airspace (NAS). The sense-and-avoid (SAA) capability has to be reliably run by the UAV in any critical situations as well. The SAA task is solved via an electro optical sensor-processor module. This module is capable of running complex processing algorithms thus it can estimate the relative pose of the camera easily. These estimates can be used in critical situations when the on-board IMU fails or can be utilized by a fused IMU/visual sensor system in order to achieve better accuracy in camera based task. The exhaustive investigation of the proposed sensor fusion architecture is not complete, however, the presented results are promising and are an excellent base for the work which is currently running and will be finished in the near future.

Update Item