LUCID — Lunar Unsupervised Classification of Illumination and Darkness
2025AI / Vision
Self supervised U-Net model for pixel level shadow illumination segmentation in Chandrayaan-2 OHRC permanently shadowed region imagery.

PythonU-NetPyTorch
Overview
LUCID (PSR Shadow Boundary Segmentation) learns pixel level shadow illumination boundaries in lunar permanently shadowed regions from Chandrayaan-2 Orbiter High-Resolution Camera imagery, using a self supervised U-Net.
Stack
Python, U-Net, PyTorch.