RaD-VIO: Rangefinder-aided Downward Visual-Inertial Odometry

RaD-VIO: Rangefinder-aided Downward Visual-Inertial Odometry

State-of-the-art forward facing monocular visual-inertial odometry algorithms are often brittle in practice, especially whilst dealing with initialisation and motion in directions that render the state unobservable. In such cases having a reliable complementary odometry algorithm enables robust and resilient flight. Using the common local planarity assumption, we present a fast, dense, and direct frame-to-frame visual-inertial odometry algorithm for downward facing cameras that minimises a joint cost function involving a homography based photometric cost and an IMU regularisation term. Via extensive evaluation in a variety of scenarios we demonstrate superior performance than existing state-of-the-art downward facing odometry algorithms for Micro Aerial Vehicles (MAVs).

People

Bo Fu
MS Student, 2019. Now a PhD student at the University of Michigan.
Kumar Shaurya Shankar
PhD, 2020. Now at Microsoft.
Nathan Michael
PI, 2014-2023. Now at Shield AI.

Robots

Acerodon
Acerodon