OrTeN: Neural Inverse Graphics for Joint Terrain and Line-scan Camera Trajectory Reconstruction

Key points

• Digital terrain maps (DTMs) are used for landing site selection, terrain-relative navigation, and geological analysis in planetary exploration.1
• Existing state-of-practice reconstruction techniques for creating DTMs from satellite imagery use multi-view stereo, requiring complex pipelines for dealing with camera error, changing illumination, and images with limited disparity.1
• OrTeN operates directly on calibrated line-scan images with pixel values expressed as radiance factor (I/F) and corrects both broad trajectory pose inconsistencies and high-frequency pointing jitter jointly with surface reconstruction during training.1
• OrTeN requires only calibrated line-scan images and initial camera poses and does not use any structural priors including ranging/depth measurements or initial rough DTMs. OrTeN is evaluated on four planetary imagery datasets.1

Sources (1)

• [1]OrTeN: Neural Inverse Graphics for Joint Terrain and Line-scan Camera Trajectory Reconstruction
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 16, 06:02 PM

  This work introduces Orbital Terrain Networks (OrTeN), a DTM construction method that uses neural inverse rendering to jointly optimize a textured height map and orbital camera corrections from line-scan satellite images.

  Digital terrain maps (DTMs) are used for landing site selection, terrain-relative navigation, and geological analysis in planetary exploration.

Extractive summary: sentences quoted from the sources.