Learn how to use methods for satellite image classification and for spatial modelling and analysis in urban applications.
Explore advanced techniques in satellite image classification and spatial modeling for urban applications through the lens of Geospatial Artificial Intelligence (GeoAI). This course delves into the intersection of GeoAI, Geographic Information Systems (GIS), and remote sensing (RS) technologies, offering insights into their application in urban planning.
Upon completion of this course, students will:
Knowledge and understanding
- Grasp fundamental GeoAI concepts.
- Explain the integration of GeoAI with GIS and RS technologies in urban planning.
Skills and abilities
- Apply GeoAI methods for satellite image classification.
- Utilize GeoAI techniques for spatial modeling and analysis in urban contexts.
Assessment skills and approaches
- Critically assess GeoAI processing and applications.
- Evaluate the pros and cons of various GeoAI methods.
1. Module 1: Theory (4 hours)
– Explore theories of GeoAI and its advantages over traditional AI in spatial modeling.
– Review optimization, machine learning, and simulation techniques.
2. Module 2: Satellite Image Classification (4 hours)
– Learn AI and deep learning methods for satellite image classification.
– Address questions on model creation and parameter settings.
– Engage in practical exercises using Copernicus data for urban object identification.
3. Module 3: Optimisation (4 hours)
– Focus on multi-objective optimization for solving spatial problems in urban areas.
– Participate in practical exercises using techniques such as NSGA-II for site selection.
4. Module 4: Simulation (4 hours)
– Develop spatiotemporal simulation models.
– Engage in practical exercises related to flood risk modeling in urban areas.
5. Module 5: Project (3 weeks)
– Define a project topic and work collaboratively in groups to implement GeoAI techniques.
– Choose a topic related to ongoing projects at SWECO or address a persistent issue using GeoAI.
– Present and discuss project results in a dedicated session.
This comprehensive course encourages hands-on learning with practical exercises and explores relevant data sources for each module.