3 June 2014 Micro air vehicle autonomous obstacle avoidance from stereo-vision
Author Affiliations +
We introduce a new approach for on-board autonomous obstacle avoidance for micro air vehicles flying outdoors in close proximity to structure. Our approach uses inverse-range, polar-perspective stereo-disparity maps for obstacle detection and representation, and deploys a closed-loop RRT planner that considers flight dynamics for trajectory generation. While motion planning is executed in 3D space, we reduce collision checking to a fast z-buffer-like operation in disparity space, which allows for significant speed-up compared to full 3d methods. Evaluations in simulation illustrate the robustness of our approach, whereas real world flights under tree canopy demonstrate the potential of the approach.
© (2014) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Roland Brockers, Roland Brockers, Yoshiaki Kuwata, Yoshiaki Kuwata, Stephan Weiss, Stephan Weiss, Lawrence Matthies, Lawrence Matthies, } "Micro air vehicle autonomous obstacle avoidance from stereo-vision", Proc. SPIE 9084, Unmanned Systems Technology XVI, 90840O (3 June 2014); doi: 10.1117/12.2055288; https://doi.org/10.1117/12.2055288

Back to Top