Lunar rover feature matching methods compared for better visual odometry
Evaluation of Front-End Feature Matching for Lunar Visual SLAM
Computer Vision and Pattern Recognition
Summary
Lunar rovers use cameras to understand where they are on the moon’s surface since there are no GPS satellites there. To figure out their movement, the rovers match visual features between camera images. This paper compares different ways of matching these features using real and simulated lunar images, finding that some newer learned methods work better in certain lighting and terrain conditions. The authors also show that results from simulated data don’t always predict how well these methods will work on real lunar images. Their benchmarks help identify which techniques track accurately and consistently during rover movement.
What this means in practice
- •For space mission engineers: Improve lunar rover navigation by selecting feature matchers that perform best under low-texture and challenging lighting.
- •For autonomous vehicle developers: Refine visual odometry techniques for off-road vehicles in low-contrast environments using evaluated learned detector-matchers.
Authors
Guillem Casadesus Vila, Grace Gao
Abstract
Lunar rovers will primarily rely on cameras for localization until navigation satellite services are deployed, and will continue to need them afterwards for relative navigation, hazard detection, and surface operations. Visual odometry (VO) and Simultaneous Localization and Mapping (SLAM) mostly estimate motion by matching features between frames. However, feature matchers are developed on urban, indoor and driving scenes, planetary studies compare them on orbital registration or within complete SLAM systems, and real lunar imagery with reference poses is scarce, so matching between rover frames on low-texture, harshly lit terrain, and its effect on VO, has not been compared. We benchmark eleven classical and learned detector-matcher combinations on nine synthetic, lunar-testbed, and terrestrial-analogue datasets, and a feature tracker, scoring pairwise pose accuracy against baseline and the trajectory error of the resulting VO. Under a common rotation-only metric, the SuperPoint pairings outperform SIFT with ratio-tested FLANN matching on both lunar testbeds and one analogue site, with smaller or unresolved differences on synthetic data. When rotation and translation direction are scored jointly, this SIFT pairing exceeds both SuperPoint pairings on four synthetic datasets, so synthetic results alone do not establish relative performance on real rover imagery. In stereo VO, the SuperPoint pairings are not consistently more accurate than SIFT, but only they keep tracking on all eight sequential datasets. Pairings with classical detectors depend on lighting while SuperPoint changes little: a low spotlight makes them more accurate than diffuse light on the same testbed views, consistent with more contrast and detected keypoints. SuperPoint with LightGlue is the fastest learned combination (16 ms per frame), and DINOv2 retrieval proposals still need geometric verification.