Kersapati, MI;
Grau-Bové, J;
(2023)
Geographic features recognition for heritage landscape mapping – Case study: The Banda Islands, Maluku, Indonesia.
Digital Applications in Archaeology and Cultural Heritage
, 28
, Article e00262. 10.1016/j.daach.2023.e00262.
Preview |
Text
1-s2.0-S2212054823000073-main.pdf - Published Version Download (12MB) | Preview |
Abstract
This study examines methods of geographic features recognition from historic maps using CNN and OBIA. These two methods are compared to reveal which one is most suitable to be applied to the historic maps dataset of the Banda Islands, Indonesia. The characteristics of cartographic images become the main challenge in this study. The geographic features are divided into buildings, coastline, and fortress. The results show that CNN is superior to OBIA in terms of statistical performance. Buildings and coastline give excellent results for CNN analysis, while fortress is harder to be interpreted by the model. On the other hand, OBIA reveals a very satisfying result is very depending on the maps’ scales. In the aspect of technical procedure, OBIA offers easier steps in pre-processing, in-process and post-processing/finalisation which can be an advantage for a wide range of users over CNN.
| Type: | Article |
|---|---|
| Title: | Geographic features recognition for heritage landscape mapping – Case study: The Banda Islands, Maluku, Indonesia |
| Open access status: | An open access version is available from UCL Discovery |
| DOI: | 10.1016/j.daach.2023.e00262 |
| Publisher version: | https://doi.org/10.1016/j.daach.2023.e00262 |
| Language: | English |
| Additional information: | © 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). |
| Keywords: | CNN, Computer vision, Historic maps, Machine learning, OBIA |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of the Built Environment UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of the Built Environment > Bartlett School Env, Energy and Resources |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10175686 |
Archive Staff Only
 |
View Item |
