Skip to content

Latest commit

 

History

History
18 lines (12 loc) · 835 Bytes

README.md

File metadata and controls

18 lines (12 loc) · 835 Bytes

FastAI4DEM

FastAI templates for Super-resolution of Digital Elevation Models.

  • SRCNN
  • SRCNN with ResBlock
  • EPSCN

These are only templates, no better results could be achieved than with a simple bicubic upsampling. If you achieve any proper results please get in contact.

Have a look at https://opendem.info/superResolution.html for detailed background information.

Based on the implementations of benymd GitHub.

The Notebook is running on Google Colab with data storage on Google Drive. Of course this could be adapted to local resources.

Download the training and validation data of the austrian DTM here. License: Data by - Geoland.at (2019) - CC BY 4.0 (CC BY 4.0)