MANGO: A Global Single-Date Paired Dataset for Mangrove Segmentation

2026-01-20

AI Summary

Existing mangrove remote sensing datasets generally lack global coverage, single-date image–mask pairs, and public accessibility, which hinders the application of deep learning in mangrove monitoring. To address this gap, this work introduces MANGO, the first publicly available benchmark dataset comprising 42,703 single-date Sentinel-2 image–mangrove mask pairs spanning 1 24 countries. Leveraging an object detection–driven strategy and a pixel-level coordinate referencing mechanism, the dataset automatically aligns optimal 2020 Sentinel-2 observations with annual mangrove extent maps. Furthermore, under a mutually exclusive national partitioning scheme, the study establishes a multi-model semantic segmentation performance benchmark. This effort provides a foundational data and methodological framework for scalable, global-scale mangrove monitoring.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Machine Learning: Multi-instance/Multi-view Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Mangroves are critical for climate-change mitigation, requiring reliable monitoring for effective conservation. While deep learning has emerged as a powerful tool for mangrove detection, its progress is hindered by the limitations of existing datasets. In particular, many resources provide only annual map products without curated single-date image-mask pairs, limited to specific regions rather than global coverage, or remain inaccessible to the public. To address these challenges, we introduce MANGO, a large-scale global dataset comprising 42,703 labeled image-mask pairs across 124 countries. To construct this dataset, we retrieve all available Sentinel-2 imagery within the year 2020 for mangrove regions and select the best single-date observations that align with the mangrove annual mask. This selection is performed using a target detection-driven approach that leverages pixel-wise coordinate references to ensure adaptive and representative image-mask pairings. We also provide a benchmark across diverse semantic segmentation architectures under a country-disjoint split, establishing a foundation for scalable and reliable global mangrove monitoring.

Problem

• mangrove segmentation
• single-date image-mask pairs
• global dataset
• remote sensing
• deep learning

Innovation

• mangrove segmentation
• single-date image-mask pairing
• target detection-driven selection
• global remote sensing dataset
• Sentinel-2