Tools
Tools
Vucinic, Luka 
ORCID: https://orcid.org/0000-0002-9370-3308, Ajia, Fatima, Suthar, Shankar Lal and Arambegedara, Charunika
(2024)
Advancing karst exploration with remote sensing and artificial intelligence: a framework for earth and planetary karst systems.
In: UNSPECIFIED.
Abstract
Remote sensing technologies have revolutionised the study of Earth’s surface, including the exploration of karst terrains, particularly in remote or inaccessible areas. This study utilises high-resolution satellite imagery and digital elevation models (DEMs) to identify and map karst landforms, with a special focus on potential subsurface access points. We are exploring various advanced image analysis techniques, including object-based image analysis and decision rule sets, to efficiently detect and classify karst features. To date, we have focused on the Maganik karst, a hard-to-reach limestone plateau in Montenegro, known for its complexity, dramatic vertical relief, and presence of multiple subsurface access points and caves, including Iron Deep, the deepest cave in the country, which reaches depths of over 1,000 metres. While our approaches can be described as semi-automated, the database developed from this study will serve as the foundation for training artificial intelligence (AI) models to autonomously detect karst features across various terrains. This approach has implications not only for Earth-based research but also for planetary exploration, where karst and karst-like processes may occur. As caves and other karst features formed through dissolution are expected in various extraterrestrial environments, including Mars and Titan, we propose a scalable framework for detecting karst features on our planet that could significantly advance our understanding of karstic landscapes beyond Earth.
| Item Type: | Conference or Workshop Item (Other) |
|---|---|
| Faculty \ School: | Faculty of Science > School of Environmental Sciences |
| Depositing User: | LivePure Connector |
| Date Deposited: | 20 Aug 2026 13:38 |
| Last Modified: | 20 Aug 2026 13:38 |
| URI: | https://ueaeprints.uea.ac.uk/id/eprint/104278 |
| DOI: |
Actions (login required)
 |
View Item |