| Autor: |
Khozouie, Nasim, Ansarifard, Zahra |
| Předmět: |
|
| Zdroj: |
International Journal of Nonlinear Analysis & Applications; Jan2025, Vol. 16 Issue 1, p319-328, 10p |
| Abstrakt: |
Accurate and automated classification of satellite images is crucial for disaster management and monitoring climate change. It is important not only to identify objects and entities in satellite images but also to reason about them and respond to queries from human operators to guide decision-making processes. Recent studies by climate researchers indicate that several areas in Shiraz are at risk of flooding. The Shiraz flood scenario is a real possibility. We have defined a disaster scenario in which the central part of Shiraz is mostly covered by water. The main objective of this research is to display the geometry of regions on a map, allowing for questions related to topology and neighborhood to be answered. Our research presents a framework for transferring satellite image data to an interactive map that is ready for mining. To obtain a searchable map from satellite data, a CNN classifier sensitive to image features is used to label regions. The framework's capabilities in terms of route connectivity are demonstrated. The features are represented in an ontology that extends the existing GeoSPARQL ontology, allowing the system to automatically search for classified regions based on specific environmental criteria. We have demonstrated how semantically enriching the representation of regions in OntoCity can improve search time, including region revision and co-routing, by enabling the system to automatically find options for regions. The SemCityMap framework can now serve as a tool for better decision-making and situational awareness. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
