← Back to NASA Technology Projects
Completed TRL 2 (started at 2, targeting 3)
Stereo vision has long been a critical technology for Mars rovers, enabling the reconstruction of 3D terrain geometry and facilitating obstacle avoidance. Since the Mars Exploration Rovers mission in the early 2000s, the foundational algorithm for stereo processing has remained largely unchanged. While demonstrating reliability and robustness, it possesses limitations, such as the creation of "holes" under unfavorable lighting conditions (e.g., the opposition effect). Concurrently, available onboard compute resources are projected to increase dramatically with the advent of space-qualified high-performance computing. Meanwhile, substantial advancements have been made in terrestrial autonomous driving, primarily propelled by deep learning algorithms. This project endeavors to leverage the latest advancements in onboard computation and algorithms to enhance the robustness of onboard stereo processing, particularly in suboptimal lighting conditions and on featureless terrains. We are currently evaluating multiple learning-based and non-learning-based techniques using over 10,000 image pairs from the Perseverance rover to compare their accuracy, robustness, and computational requirements. Additionally, we are investigating near-term improvements that can be implemented on Perseverance through parameter modifications.
The primary benefit out of this project is the creation of new stereo processing approach, including downsampling, pre- and post-filtering, and stereo correlation, that is suitable for future low-cost mobility missions on Mars. The new technology will be applicable for both ground and aerial mobility. Furthermore, the project will aim to produce short-term benefits by improving Perservance’s stereo performance through parameter changes. This could include a selection of different de-Bayering approach and updating the look-up table for bit downsampling. This project's primary benefit is the development of a novel stereo processing approach, encompassing downsampling, pre- and post-filtering, and stereo correlation, specifically designed for future low-cost mobility missions on Mars. This new technology will be applicable to both ground and aerial mobility. Additionally, the project aims to yield immediate benefits by enhancing Perseverance's stereo performance through parameter modifications, potentially including the selection of a different de-Bayering approach and an update to the bit downsampling look-up table.
Listed on TechPort itself — the most direct way to ask about this specific project.
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.