Vision Based Navigation

Collision-Free Autonomous Navigation

Safe and accurate navigation for autonomous trajectory tracking of quadrotors {using monocular vision} is addressed in this work. A second order Sliding Mode (2-SM) control algorithm is used to track desired trajectories, providing robustness against model uncertainties and {external} perturbations. A Lyapunov based analysis proved the closed-loop stability of the system despite the presence of unknown external perturbations. Monocular vision using PTAM (Parallel Tracking and Mapping), fused with inertial measurements are employed to estimate the vehicle's pose with respect to unstructured scenes. In addition, the distance to potential collisions is detected and computed using the sparse depth map coming also from the vision algorithm. Potential fields are employed to drive the UAV away from collisions. The proposed strategy is successfully tested in real-time experiments.

Obstacle avoidance experiment. The UAV is able to detect the obstacle and go around it to continue its path.

UAV localization w.r.t. the sparse depth map using PTAM (left). Characteristic features on the image (right).

Collision-free trajectory tracking. The aerial vehicle is capable of following the trajectory except when the obstacle is detected and a repulsive forces is induced. The desired path is a lemniscate and the wall is represented by the black rectangle.

Videos

Publications

[J4] D. Mercado, P. Castillo and R. Lozano. Sliding Mode Collision-Free Navigation for Quadrotors using Monocular Vision, Robotica, 2018, 1-17. doi:10.1017/S0263574718000516.

[C5] D. Mercado, P Castillo, R Lozano. Quadrotor's trajectory tracking control using monocular vision navigation. Unmanned Aircraft Systems (ICUAS), International Conference on, 844-850. 2015.

Google Sites

Report abuse