Refinement of the Mars 2020 Terrain Relative Navigation HiRISE DTM Mosaic using Deep Learning

Abstract

We present an enhanced High-Resolution Imaging Science Experiment (HiRISE) digital terrain model (DTM) mosaic of the Mars 2020 landing site at Jezero crater. Our DTM mosaic is based on the monocular depth estimation network called MADNet (Tao et al., 2021), which significantly improves the effective resolution and eliminates artifacts. To generate this product, we use the publicly available Mars 2020 TRN (Terrain Relative Navigation) orthoimage and DTM mosaics, provided by the United States Geological Survey (USGS), as inputs. The resultant HiRISE MADNet DTM mosaic is strictly co-aligned with the original USGS DTM and orthoimage mosaics, which are already co-aligned with the USGS Context Camera (CTX) based DTM and orthoimage mosaics, as well as the ESA/DLR/FUB High Resolution Stereo Camera (HRSC) level 5 DTM and orthoimage mosaics that cover the same area.