Machine learning digitization of historical land use

Project Details

Abstract

Machine learning techniques and the availability of georeferenced historical maps offer new opportunities to digitize historical land use. A cooperation between several Flemish authorities (INBO, department of agriculture and fisheries, Digital Flanders, VITO) aimed to do so for 3 or 4 historical maps, and analyse land use changes for a time span of approximately 200 years.
StatusActive
Effective start/end date1/04/2231/03/23

Free keywords

  • historicalmaps
  • landusechanges

Thematic List 2020

  • Soil & air
  • Nature & society