Poster on Extraction of Kenya's Historical Road Network from Topographic Maps using Deep Learning Techniques

This page lists all metadata that was entered for this dataset. You can download the dataset.

Feature
Citation
Citation Options
Identification
Title:Main Title: Poster on Extraction of Kenya's Historical Road Network from Topographic Maps using Deep Learning Techniques
Description:Abstract: This is a poster about the extraction of Kenya’s historical road network from topographical maps presented at the DGPF annual meeting 2024 in Remagen, Germany. Kenya’s road network is crucial for environmental and socio-economic development. High-quality road data is vital for analyzing its impact on various factors. However, Kenya, like many sub-Saharan countries, faces challenges in digital road data availability due to high data acquisition costs. To address this, we extracted Kenya’s road network from over 500 historical maps using deep learning. We digitized, georeferenced, and classified the maps, identifying over 20 road symbols. A Python script and ArcGIS Pro’s Multi-Task Road Extractor were used for classification. The result is a detailed dataset of Kenya’s roads from the 1950s to 1970s. Despite some classification issues, the method is promising for national-scale road delineation and can be applied to other countries’ historical maps.
Related Resource:Is Supplemented By https://www.trr228db.uni-koeln.de/search/view.php?dataID=492 (URL)
Responsible Party
Creators:Tanja Kramm (Author), Nicodemus Nyamari (Author), Vincent Moseti (Author), David M. Anderson (Author), Christina Bogner (Author), Georg Bareth (Author)
Funding Reference:Deutsche Forschungsgemeinschaft (DFG): CRC/TRR 228: Future Rural Africa: Future-making and social-ecological transformation
Publisher:TRR228 Database (TRR228DB)
Publication Year:2024
Topic
TRR228 Topic:Infrastructure
Related Subprojects:Z2, A5, A2
Subjects:Keywords: Roads, Road Network, Geodata, Historical Data, Infrastructure, Historical Maps
Geogr. Information Topic:Geoscientific Information
File Details
Filename:Poster_DGPF2024.pdf
Data Type:Text - Poster
File Size:2.3 MB
Date:Available: 13.03.2024
Mime Type:application/pdf
Data Format:PDF
Language:English
Status:Completed
Constraints
Download Permission:Free
General Access and Use Conditions:According to the TRR228DB data policy agreement.
Access Limitations:According to the TRR228DB data policy agreement.
Licence:[TRR228DB] Data policy agreement
Geographic
Specific Information - Presentation
Presenter:Tanja Kramm
Presentation Date:13th of March, 2024
Presentation Type:Poster
Event:Jahrestagung 2024 in Remagen: Stadt, Land, Fluss - Daten vernetzen
Event Type:Conference
Event Location:Remagen, Germany
Event Duration:13th of March, 2024 - 14th of March, 2024
Event Website:https://www.dgpf.de/con.html
Metadata Details
Metadata Creator:Tanja Kramm
Metadata Created:19.04.2024
Metadata Last Updated:19.04.2024
Subproject:Z2
Funding Phase:2
Metadata Language:English
Metadata Version:V50
Metadata Export
Metadata Schema:
Dataset Statistics
Page Visits:23
Metadata Downloads:0
Dataset Downloads:0
Dataset Activity
Feature