Intelligent Perception for Autonomous Characterization and Relative Navigation of Small Bodies

The goal of this research grant is to develop novel vision-based methods to autonomously estimate a robotic spacecraft’s pose relative to a target small body while simultaneously reconstructing and characterizing the body's surface to reduce reliance on humans-in-the-loop and to provide an autonomous alternative to state-of-the-practice small body shape reconstruction methods. The proposed methodology is comprised of four key ingredients, followed by a rigorous validation plan. First, data-driven methods that leverage deep convolution neural networks (CNNs) will be employed to jointly learn robust keypoint detectors and feature descriptors for improved data association in the front-end of the proposed system. Second, optimal information fusion between different sensor modalities and application-specific constraints will be accomplished via graph-based smoothing in the back-end of the system to concurrently estimate the pose of the spacecraft and a map of the body. Third, photometric stereo techniques will be factored into the proposed system to estimate surface gradients and albedos to provide detailed topological information to aid in 3D reconstruction and characterization of the body. Fourth, efficient graph cut optimization methods and an adaptive space discretization will be employed to construct a mesh of the body using the spatial and topological information. Finally, the proposed approach will be validated using real imagery from past small body missions as well as experimentally tested on the ASTROS platform at Georgia Tech.

This is early/mid-stage (TRL 2) — the most realistic path in is NASA SBIR/STTR, which funds small businesses and research institutions to develop technology aligned with NASA's needs (equity-free, phased funding). Check whether a current SBIR/STTR solicitation topic overlaps with this project's technology area, or contact the project directly (above) to ask.

None of these are guaranteed paths for this specific project — TechPort itself doesn't have an "apply" button. Reaching out to the contact(s) above with a specific question is usually the fastest way to find out what's actually open.