Context-Aware Slum Mapping in Sub-Saharan Africa Using Sentinel-1 Texture and Local Climate Zones

2026-07-08 · Source: Computer Vision and Pattern Recognition · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Expert, quick

Summary

A new context-aware Optical-SAR framework significantly improves informal settlement mapping in Sub-Saharan African cities, addressing the challenge of distinguishing LCZ 7 informal settlements from spectrally similar LCZ 3 formal compact low-rise areas. This reproducible framework integrates Sentinel-2 spectral features with Sentinel-1 structural information, employing a three-tier SAR strategy including calibrated backscatter, GLCM textures, and a physics-guided feature designed for high structural disorder. Evaluated in Nairobi and Eldoret, Kenya, the Optical-SAR model achieved overall accuracies of 0.816 (dry) and 0.807 (wet), substantially outperforming the WUDAPT baseline (OA 0.704) and reducing critical LCZ 3 - LCZ 7 confusion to 7%. SAR textures were identified as the dominant performance driver, stabilizing informal settlement mapping across seasons where optical-only methods varied. Cross-city transfer, however, requires local adaptation strategies.

Key takeaway

For Computer Vision Engineers developing urban morphology mapping solutions in data-scarce regions, you should prioritize integrating Sentinel-1 SAR-derived textures with optical imagery. This approach significantly reduces confusion between informal settlements (LCZ 7) and formal low-rise areas (LCZ 3), achieving consistent accuracy across seasons where optical-only methods falter. Consider developing local adaptation strategies for cross-city deployment to maximize framework effectiveness.

Key insights

Integrating Sentinel-1 SAR textures with optical data significantly improves informal settlement mapping in Sub-Saharan Africa, stabilizing results across seasons.

Principles

• SAR textures are crucial for distinguishing urban morphology.
• Context-aware frameworks enhance mapping accuracy.
• Seasonal stability is achievable with SAR integration.

Method

The framework uses Sentinel-2 spectral features and Sentinel-1 SAR (backscatter, GLCM textures, physics-guided feature) within an adapted Local Climate Zone taxonomy for informal settlement delineation.

In practice

• Use Sentinel-1 SAR textures for urban morphology mapping.
• Apply LCZ taxonomy for informal settlement classification.
• Combine optical and SAR data for robust seasonal mapping.

Topics

• Informal Settlement Mapping
• Sentinel-1 SAR
• Sentinel-2 Optical
• Local Climate Zones
• Urban Morphology
• Sub-Saharan Africa

Best for: AI Scientist, Computer Vision Engineer, Research Scientist

Related on AIssential

• Heterogeneous SAR-optical fusion for near-real-time land use and land cover mapping under cloud contamination: A novel framework and global benchmark dataset — CloudLULC-Net fuses SAR and cloud-contaminated optical data for robust LULC mapping, overcoming traditional cloud limitations.
• MapSR: Prompt-Driven Land Cover Map Super-Resolution via Vision Foundation Models — MapSR uses prompt-driven vision foundation models for efficient, training-free land-cover super-resolution with minimal annotation.
• JL1-CC&QA: Extending the JL1-CD Benchmark with Change Captioning and Question Answering — Remote sensing change detection benefits from semantic layers like captioning and QA to understand "what" and "why" changes occur.
• Contrasting Local and Global Modeling with Machine Learning and Satellite Data: A Case Study Estimating Tree Canopy Height in African Savannas — Local SatML models can outperform global or fine-tuned global models for specific regional tasks.
• DiffuSAM: Diffusion Guided Zero-Shot Object Grounding for Remote Sensing Imagery — DiffuSAM combines diffusion models with segmentation models for superior zero-shot object grounding in remote sensing.
• Robust Multispectral Semantic Segmentation under Missing or Full Modalities via Structured Latent Projection — Structured latent projection in CBC-SLP preserves modality-specific information, enhancing multimodal semantic segmentation robustness.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Computer Vision and Pattern Recognition.