UrbanOccupationsOETR_Deutsche_Heereskarte_200_Turkey
Description
Data: This section of the historical DHK 200 Turkey map covers a large area of around 150,000 square km in Turkey, including the regions of Balıkesir, Bursa, Eskişehir, Ankara, and Yozgat. The DHK 200 data consist of 65,000 x 18,000 pixels with a dpi of 300 and has a raster size of 17 x 17 meters. The map has a radiometric resolution of 8-bit and is in the RGB space. Dataset: The dataset constructed from the DHK 200 data consist of 256 x 256 dual patches (map and the corresponding ground truth) and contain 5 classes; stabilized roads (German: Fahrstraße, Turkish: şose), railways (German: Eisenbahn, Turkish: demiryolu), footpaths (German: Fußweg, Turkish: patika), cart roads (German: Fahrweg/Karawanenweg, Turkish: araba yolu), inferior roads (German: Saumweg, Turkish: adi yol). If you would like to use the dataset below in further publication, please use the credentials specified below: Ekim, Burak, Elif Sertel, and M. E. Kabadayı 2021. “Automatic Road Extraction from Historical Maps Using Deep Learning Techniques: A Regional Case Study of Turkey in a German World War II Map” ISPRS International Journal of Geo-Information 10, no. 8: 492. https://doi.org/10.3390/ijgi10080492
Resources
| Name |
Format |
Description |
Link |
|
0 |
|
http://data.europa.eu/88u/dataset/oai-zenodo-org-7072623 |
|
0 |
|
http://data.europa.eu/88u/dataset/oai-zenodo-org-7072623 |
Tags
- convolutional-neural-networks;-road-classification;-segmentation;-deep-learning;-fully-convolutional-networks;-historical-maps