This repository contains the code for the paper "Automatic Road Extraction from Historical Maps using Deep Learning Techniques: A Regional Case Study of Turkey in a German World War II map"
Modify the paths in the params.py script and run the main.ipynb notebook. It is also possible to run evaluation and inference parts of the notebook alone to reproduce the results stated in the paper. To do so, use the link below to get the data and the weights.
The weights and the splitted dataset can be found at the following link:
https://drive.google.com/drive/folders/1nTO_ux02-vh3p4R6fmzVF_8ZxfHvPkJb?usp=sharing
If you want to run the code in your own data, you can accordingly change the params.py file (e.g., data and ground truth paths, num_class) and tune the hyperparameters.
The code was implemented in Python(3.8) and PyTroch(1.14.0) on Windows OS. The segmentation models pytorh library is used as a baseline for implementation. Apart from main data science libraries, RS-specific libraries such as GDAL, rasterio, and tifffile are also required.
Please kindly cite the paper below if this code is useful and helpful for your research.
TBD
If you encounter bugs while using this code, please do not hesitate to contact us.
- Burak Ekim: ekim19@itu.edu.tr