Mapping the Green Urban: A Comprehensive Review of Materials and Learning Methods for Green Infrastructure Mapping

Area of Science:

• Environmental Science
• Urban Planning
• Remote Sensing

Background:

• Green infrastructure (GI) is vital for sustainable urban development.
• Effective mapping of GI requires understanding materials and advanced methods.
• Current mapping approaches need detailed analysis of sensors, data, and algorithms.

Purpose of the Study:

• To review the state-of-the-art in green infrastructure mapping.
• To analyze the application of machine learning and deep learning in GI mapping.
• To present a novel methodology for automated verification in GI mapping.

Main Methods:

• Systematic literature review using PRISMA guidelines.
• Natural language processing and large language models for automated screening.
• Analysis of 55 papers focusing on GI mapping materials and learning methods.

Main Results:

• A significant shift towards deep learning methods in GI mapping (33 papers) compared to machine learning (22 papers).
• Identification of various sensors and image data utilized in GI mapping.
• Demonstration of a novel automated verification methodology for GI mapping.

Conclusions:

• Deep learning is becoming dominant in green infrastructure mapping.
• Automated verification methods show promise for enhancing GI mapping accuracy.
• Further research is needed to refine automated verification techniques and address current limitations.