A 1.2 Billion Pixel Human-Labeled Dataset for Data-Driven Classification of Coastal Environments
Abstract The world’s coastlines are spatially highly variable, coupled-human-natural systems that comprise a nested hierarchy of component landforms, ecosystems, and human interventions, each interacting over a range of space and time scales. Understanding and predicting coastline dynamics necessita...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Portfolio
2023-01-01
|
Series: | Scientific Data |
Online Access: | https://doi.org/10.1038/s41597-023-01929-2 |
_version_ | 1797946078853595136 |
---|---|
author | Daniel Buscombe Phillipe Wernette Sharon Fitzpatrick Jaycee Favela Evan B. Goldstein Nicholas M. Enwright |
author_facet | Daniel Buscombe Phillipe Wernette Sharon Fitzpatrick Jaycee Favela Evan B. Goldstein Nicholas M. Enwright |
author_sort | Daniel Buscombe |
collection | DOAJ |
description | Abstract The world’s coastlines are spatially highly variable, coupled-human-natural systems that comprise a nested hierarchy of component landforms, ecosystems, and human interventions, each interacting over a range of space and time scales. Understanding and predicting coastline dynamics necessitates frequent observation from imaging sensors on remote sensing platforms. Machine Learning models that carry out supervised (i.e., human-guided) pixel-based classification, or image segmentation, have transformative applications in spatio-temporal mapping of dynamic environments, including transient coastal landforms, sediments, habitats, waterbodies, and water flows. However, these models require large and well-documented training and testing datasets consisting of labeled imagery. We describe “Coast Train,” a multi-labeler dataset of orthomosaic and satellite images of coastal environments and corresponding labels. These data include imagery that are diverse in space and time, and contain 1.2 billion labeled pixels, representing over 3.6 million hectares. We use a human-in-the-loop tool especially designed for rapid and reproducible Earth surface image segmentation. Our approach permits image labeling by multiple labelers, in turn enabling quantification of pixel-level agreement over individual and collections of images. |
first_indexed | 2024-04-10T21:05:12Z |
format | Article |
id | doaj.art-45d678cc93a240f8a63e5767197bb636 |
institution | Directory Open Access Journal |
issn | 2052-4463 |
language | English |
last_indexed | 2024-04-10T21:05:12Z |
publishDate | 2023-01-01 |
publisher | Nature Portfolio |
record_format | Article |
series | Scientific Data |
spelling | doaj.art-45d678cc93a240f8a63e5767197bb6362023-01-22T12:04:16ZengNature PortfolioScientific Data2052-44632023-01-0110111810.1038/s41597-023-01929-2A 1.2 Billion Pixel Human-Labeled Dataset for Data-Driven Classification of Coastal EnvironmentsDaniel Buscombe0Phillipe Wernette1Sharon Fitzpatrick2Jaycee Favela3Evan B. Goldstein4Nicholas M. Enwright5Contractor, U.S. Geological Survey Pacific Coastal and Marine Science CenterU.S. Geological Survey Pacific Coastal and Marine Science CenterContractor, U.S. Geological Survey Pacific Coastal and Marine Science CenterContractor, U.S. Geological Survey Pacific Coastal and Marine Science CenterDepartment of Geography, Environment, and Sustainability, University of North Carolina at GreensboroU.S. Geological Survey Wetland and Aquatic Research CenterAbstract The world’s coastlines are spatially highly variable, coupled-human-natural systems that comprise a nested hierarchy of component landforms, ecosystems, and human interventions, each interacting over a range of space and time scales. Understanding and predicting coastline dynamics necessitates frequent observation from imaging sensors on remote sensing platforms. Machine Learning models that carry out supervised (i.e., human-guided) pixel-based classification, or image segmentation, have transformative applications in spatio-temporal mapping of dynamic environments, including transient coastal landforms, sediments, habitats, waterbodies, and water flows. However, these models require large and well-documented training and testing datasets consisting of labeled imagery. We describe “Coast Train,” a multi-labeler dataset of orthomosaic and satellite images of coastal environments and corresponding labels. These data include imagery that are diverse in space and time, and contain 1.2 billion labeled pixels, representing over 3.6 million hectares. We use a human-in-the-loop tool especially designed for rapid and reproducible Earth surface image segmentation. Our approach permits image labeling by multiple labelers, in turn enabling quantification of pixel-level agreement over individual and collections of images.https://doi.org/10.1038/s41597-023-01929-2 |
spellingShingle | Daniel Buscombe Phillipe Wernette Sharon Fitzpatrick Jaycee Favela Evan B. Goldstein Nicholas M. Enwright A 1.2 Billion Pixel Human-Labeled Dataset for Data-Driven Classification of Coastal Environments Scientific Data |
title | A 1.2 Billion Pixel Human-Labeled Dataset for Data-Driven Classification of Coastal Environments |
title_full | A 1.2 Billion Pixel Human-Labeled Dataset for Data-Driven Classification of Coastal Environments |
title_fullStr | A 1.2 Billion Pixel Human-Labeled Dataset for Data-Driven Classification of Coastal Environments |
title_full_unstemmed | A 1.2 Billion Pixel Human-Labeled Dataset for Data-Driven Classification of Coastal Environments |
title_short | A 1.2 Billion Pixel Human-Labeled Dataset for Data-Driven Classification of Coastal Environments |
title_sort | 1 2 billion pixel human labeled dataset for data driven classification of coastal environments |
url | https://doi.org/10.1038/s41597-023-01929-2 |
work_keys_str_mv | AT danielbuscombe a12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT phillipewernette a12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT sharonfitzpatrick a12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT jayceefavela a12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT evanbgoldstein a12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT nicholasmenwright a12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT danielbuscombe 12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT phillipewernette 12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT sharonfitzpatrick 12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT jayceefavela 12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT evanbgoldstein 12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments AT nicholasmenwright 12billionpixelhumanlabeleddatasetfordatadrivenclassificationofcoastalenvironments |