Deep learning model to classify flood inundated areas in multispectral satellite imagery. A brief summary of the item is not available. Add a brief summary about the item.
Deep learning package
by
Item created: Jan 3, 2024 Item updated: Jan 2, 2025 Number of downloads: 3,112
Description
For
flood monitoring and management, including disaster response in the
aftermath of a flood, identifying submerged areas is a challenge for
emergency responders and policymakers. The increase in accessibility to
earth observation data and the continuous evolution of deep learning
methods is enabling an efficient way to automate flood monitoring and
management. Flood segmentation deep learning models emerge as an
important tool in helping with the task of precisely identifying and
delineating flood-affected regions from satellite imagery.
The Prithvi-100M-sen1floods11 has been developed by NASA and IBM by finetuning their foundation model for earth observation - Prithvi-100m, using Sen1Floods11 dataset. Use this model to automate the process of segmenting flood extents in multispectral satellite imagery.
Using the model
Follow the guide to
use the model. Before using this model, ensure that the supported deep
learning libraries are installed. For more details, check Deep Learning Libraries Installer for ArcGIS.
Fine-tuning the model
This model can be fine-tuned using the ArcGIS API for Python. Follow the guide to fine-tune this model.
Input
Raster, mosaic dataset, or image service. These must be a 6-band composite raster derived from either Harmonized Landsat 8 (HLSL30) or Harmonized Sentinel 2 (HLSS30).
The model can also be used with level-2 products of Sentinel-2 and
Landsat-8, yet it performs most effectively with HLSL30 and HLSS30.
The composite raster should contain the following 6 bands: Blue, Green, Red, Narrow NIR, SWIR, SWIR 2.
Band numbers for the above mentioned bands are:
- For HLSS30 and Sentinel-2: Band2, Band3, Band4, Band8A, Band11, Band12
- For HLSL30 and Landsat 8: Band2, Band3, Band4, Band5, Band6, Band7
Output
Classified raster with 3 classes (no water, water/flood, and no data/clouds).
Applicable geographies
This model is expected to work well across the globe.
Model architecture
This model packages IBM and NASA's Prithvi-100M-sen1floods11 model and uses a self-supervised encoder developed with a ViT architecture and Masked AutoEncoder (MAE) learning strategy.
Accuracy metrics
This model has a mean intersection over union of 0.88 percent and mean accuracy of 94.37 percent.
Training data
This model finetunes the pretrained Prithvi-100m model to segment the extent of flooded area on multispectral satellite images from the Sen1Floods11 dataset.
Sample results
Here are a few results from the model.
Citations
Bonafilia, D., Tellman, B., Anderson, T., Issenberg, E. 2020. Sen1Floods11: a georeferenced dataset
to train and test deep learning flood algorithms for Sentinel-1. The IEEE/CVF Conference on
Computer Vision and Pattern Recognition (CVPR) Workshops, 2020, pp. 210-211.
An in-depth description of the item is not available.
Layers
Tools
Tables
Basemap
Project Contents:
Solution Contents
Contents
Layers
Screenshots
Terms of Use
No special restrictions or limitations on using the item's content have been provided.
Details
Dashboard views: Desktop
Source:
Creating data in:
Published as:
Other Views:
Dependent items in the recycle bin
Applicable: 2d
Size: 1,080.647 MB
ID: 29dc90c33daf402caa9293c2088d1057
Image Count: 0
Image Properties
Layer Drawing
Using tiles from a cache
Dynamically from data
Share
Owner
Folder
Categories
This item has not been categorized.
Credits (Attribution)
No acknowledgements.IBM, NASA
Comments (0)