Neural 3D Reconstruction of Planetary Surfaces from Descent-Phase Wide-Angle Imagery

2026-04-14

AI Summary

This work addresses the challenge of low parallax in wide-field-of-view descent imagery caused by strong radial distortion and near-nadir viewing during planetary landings. To tackle this issue, the study introduces neural reconstruction into planetary descent imaging for the first time, proposing an explicit neural elevation field–based method for 3D terrain modeling. The approach incorporates prior knowledge that planetary surfaces are generally continuous, smooth, and free of floating objects, and it is specifically optimized for the imaging characteristics of wide-field cameras. Experimental results on high-fidelity simulated lunar and Martian descent image sequences demonstrate that the proposed method significantly improves spatial coverage compared to conventional multi-view stereo techniques while maintaining high accuracy in elevation estimation.

Technology Category

• Computer Vision: 3D Computer Vision
• Intelligent Robots: Localization, Mapping, and Navigation
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Models for Web evolution

Abstract

Digital elevation modeling of planetary surfaces is essential for studying past and ongoing geological processes. Wide-angle imagery acquired during spacecraft descent promises to offer a low-cost option for high-resolution terrain reconstruction. However, accurate 3D reconstruction from such imagery is challenging due to strong radial distortion and limited parallax from vertically descending, predominantly nadir-facing cameras. Conventional multi-view stereo exhibits limited depth range and reduced fidelity under these conditions and also lacks domain-specific priors. We present the first study of modern neural reconstruction methods for planetary descent imaging. We also develop a novel approach that incorporates an explicit neural height field representation, which provides a strong prior since planetary surfaces are generally continuous, smooth, solid, and free from floating objects. This study demonstrates that neural approaches offer a strong and competitive alternative to traditional multi-view stereo (MVS) methods. Experiments on simulated descent sequences over high-fidelity lunar and Mars terrains demonstrate that the proposed approach achieves increased spatial coverage while maintaining satisfactory estimation accuracy.

Problem

• Neural 3D Reconstruction
• Planetary Surfaces
• Wide-Angle Imagery
• Digital Elevation Modeling
• Descent-Phase Imaging

Innovation

• neural 3D reconstruction
• planetary surfaces
• descent-phase imagery
• neural height field
• multi-view stereo