Skip to content

nalemadi/RSGeoDisasterNet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 

Repository files navigation

RSGeoDisasterNet

This repository is initialized to survey "Remote Sensing Datasets for Natural Disaster".

Note

Additional related content will be added soon!

Motivation

Remote sensing has become a cornerstone technology for monitoring, understanding, and responding to natural disasters. Advances in satellite and airborne imaging, combined with artificial intelligence, have enabled large-scale disaster detection, damage assessment, and change analysis across diverse geographic regions. In parallel, a growing number of public datasets have been released to support research on disasters such as floods, earthquakes, wildfires, hurricanes, and landslides. Despite this rapid growth, the dataset landscape remains fragmented. Existing datasets vary widely in sensor modality, spatial resolution, temporal coverage, annotation granularity, and task definition. Selecting an appropriate dataset for a given research or operational objective is therefore non-trivial, particularly for AI-based methods that rely on consistent labeling schemes and well-aligned pre- and post-event imagery.

To the best of our knowledge, no prior work has systematically surveyed remote sensing datasets for natural disasters across sensing modalities, disaster types, annotation strategies, and task formulations. This study presents a systematic survey of publicly available remote sensing datasets for natural disasters. Our goals are to:

  • Consolidate and systematically analyze existing datasets
  • Critically assess their limitations
  • Identify structural gaps to inform future dataset curation and benchmarking efforts. We introduce a structured taxonomy and evaluation framework, consolidate key dataset characteristics, identify coverage gaps, provide guidance to support informed dataset selection, and present a roadmap for future research.

This effort is particularly timely given the rapid shift toward large-scale pre-trained remote sensing foundation models, whose performance and generalization critically depend on diverse, well-annotated, and temporally aligned disaster datasets.

Scope and Dataset Collection

TBA

Dataset
xBD/ xView2: Assess Building Damage
DisasterM3
Sen1Floods11
Landslide4Sense
STURM-Flood
Hephaestus
BRIGHT
MMFlood
...

Dataset Evaluation Framework

TBA

Challenges and Future Directions

TBA

About

This repository is initialized to survey "Remote Sensing Datasets for Natural Disaster"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors