LunarDepthNet: Generation of Digital Elevation Models using Deep Learning and Monocular Satellite Images

arXiv cs.CV / 4/28/2026

📰 NewsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper introduces LunarDepthNet, a deep learning approach to generate lunar digital elevation models (DEMs) directly from monocular satellite images, addressing the lack of detailed elevation data on the Moon.
LunarDepthNet uses a UNet-based architecture with an EfficientNet encoder and custom components designed to learn the relationship between surface lighting/shadows and true elevation.
The method is trained and evaluated on Chandrayaan-2 Terrain Mapping Camera (TMC) images paired with corresponding digital terrain models (DTMs), using a combined loss to preserve terrain detail while ensuring smoothness.
Validation shows stable loss convergence of 12%, and testing results report mean nRMSE of 0.437 and MAE of 4.5 meters, indicating the model can produce reliable elevation maps when stereo imagery is unavailable.

Abstract

Recent times have seen an increase in demand of high quality Digital Elevation Models (DEMs) for the lunar surface, because they are highly important for studying the moon and planning future missions. However, there is an evident lack of detailed elevation data on the Moon. To overcome this limitation, this study proposes a novel deep learning method that estimates and generates a surface elevation map directly from monocular images of the surface. The dataset used comprises of the Chandrayaan-2 Terrain Mapping Camera (TMC) images with their corresponding Digital Terrain Models (DTMs). The study proposes LunarDepthNet, which comprises of a UNet architecture to generate DEMS. It incorporates an EfficientNet encoder and custom layers to correctly learn how the light shadows on the surface relate to the actual elevation values. A combined loss function was also utilized to keep the terrain details accurate and smooth. During validation, the model showed a stable loss convergence of 12%. It achieved a mean nRMSE of 0.437 and an MAE of 4.5m in the testing stage. These results prove the model can generate dependable elevation maps from single orbital images, which are quite useful in regions of the moon where stereo-images are not available.