Lunar Turing-GAN Super-Resolution Model

Objective: Create a high-resolution map of the Moon using a pipeline of Image Super-Resolution models

  1. Proposed a novel GAN-based architecture with turing test based adversaries for ensuring accurate reconstruction of craters and hills.
  2. Achieved a competitive SSIM of 0.794 while increasing image spatial resolution from 5m per pixel to 30 cm per pixel, a 16x magnification.
  3. Created a pipeline capable of tiling and super-resolving an image using Lunar T-GAN, HAT, RealESRGAN and sharpening algorithms.
  4. Developed a Lunar Atlas by correcting coordinates & stitching together individual image patches from the Chadrayaan-2 TMC payload.