Combining Stereo Vision and Inertial Navigation System for a Quad-Rotor UAV

Luis Rodolfo García Carrillo (1) , Alejandro Enrique Dzul López (2) , Rogelio Lozano (3, 1) , Claude Pégard (4, 5)

1 Heudiasyc - Heuristique et Diagnostic des Systèmes Complexes [Compiègne]
2 ITL - Instituto Tecnologico de la Laguna
3 UMI - UMI LAFMIA CINVESTAV
4 MIS - Modélisation, Information et Systèmes - UR UPJV 4290
5 UPJV - Université de Picardie Jules Verne

This paper presents the development of a quad-rotor robotic platform equipped with a visual and inertial motion estimation system. Our objective consists of developing a UAV capable of autonomously perform take-off, positioning, navigation and landing in unknown environments. In order to provide accurate estimates of the UAV position and velocity, stereo visual odometry and inertial measurements are fused using a Kalman Filter. Real-time experiments consisting on motion detection and autonomous positioning demonstrate the performance of the robotic platform.

Mots clés: UAV; Quad-rotor; Navigation; Localization; Computer vision; Stereo vision; Ego-motion; Data fusion; Aerial robotics; Autonomous robots

Domaines: Automatique

Luis Rodolfo García Carrillo, Alejandro Enrique Dzul López, Rogelio Lozano, Claude Pégard. Combining Stereo Vision and Inertial Navigation System for a Quad-Rotor UAV. Journal of Intelligent and Robotic Systems, 2012, 65 (1-4), pp.373-387. ⟨10.1007/s10846-011-9571-7⟩. ⟨hal-00937512⟩