Urban area extent extraction in spaceborne HR and VHR data using multi-resolution features

Area of Science:

• Remote Sensing
• Geographic Information Systems (GIS)
• Urban Studies

Background:

• Defining urban areas is challenging due to varied definitions and data resolutions.
• Existing models combine spectral and spatial information, leading to multiple extraction outputs from the same dataset.

Purpose of the Study:

• To propose a novel, flexible framework for urban area extent extraction from multispectral Earth Observation (EO) data.
• To develop an approach that can match multiple urban area models by computing and combining spectral and multi-scale spatial features.

Main Methods:

• Utilizing multispectral Earth Observation (EO) data.
• Computing and combining spectral and multi-scale spatial features.
• Applying logical rules to select and integrate features for matching diverse urban area models.

Main Results:

• The proposed framework successfully extracts urban area extents from High-Resolution (HR) data.
• Experimental results from Brazil and Kenya demonstrate the framework's usefulness and flexibility.
• The approach allows for multiple urban area model extractions from a single dataset.

Conclusions:

• The developed framework offers a versatile solution for urban area mapping using remote sensing.
• It effectively addresses the challenge of diverse urban area definitions by integrating spectral and spatial features.
• The framework's adaptability makes it valuable for various urban remote sensing applications.