Automated mapping of cultural heritage in Norway from airborne lidar data using faster R-CNN

Open Webpage

Øivind Due Trier, Jarle Hamar Reksten, Kristian Løseth

Published: 01 Jan 2021, Last Modified: 04 Nov 2025Int. J. Appl. Earth Obs. Geoinformation 2021EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•There is a need for detailed, accurate and complete cultural heritage mapping.•Simple Faster R-CNN was used for automated detection in airborne lidar data.•Object types were grave mounds, pitfall traps and charcoal kilns.•90 % of confirmed charcoal kilns were detected, with 38 % false positives.•38 % of confirmed grave mounds were detected, with 89 % false positives.