Mapping Relict Charcoal Hearths in New England Using Deep Convolutional Neural Networks and LiDAR Data

Advanced deep learning methods combined with regional, open access, airborne Light Detection and Ranging (LiDAR) data have great potential to study the spatial extent of historic land use features preserved under the forest canopy throughout New England, a region in the northeastern United States. M...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Ji Won Suh, Eli Anderson, William Ouimet, Katharine M. Johnson, Chandi Witharana
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	remote sensing relict charcoal hearths deep convolutional neural networks semantic segmentation anthropogenic feature detection Science Q
Acceso en línea:	https://doaj.org/article/847966d477074eab88e929d3419c6d03
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:847966d477074eab88e929d3419c6d03
record_format	dspace
spelling	oai:doaj.org-article:847966d477074eab88e929d3419c6d032021-11-25T18:54:54ZMapping Relict Charcoal Hearths in New England Using Deep Convolutional Neural Networks and LiDAR Data10.3390/rs132246302072-4292https://doaj.org/article/847966d477074eab88e929d3419c6d032021-11-01T00:00:00Zhttps://www.mdpi.com/2072-4292/13/22/4630https://doaj.org/toc/2072-4292Advanced deep learning methods combined with regional, open access, airborne Light Detection and Ranging (LiDAR) data have great potential to study the spatial extent of historic land use features preserved under the forest canopy throughout New England, a region in the northeastern United States. Mapping anthropogenic features plays a key role in understanding historic land use dynamics during the 17th to early 20th centuries, however previous studies have primarily used manual or semi-automated digitization methods, which are time consuming for broad-scale mapping. This study applies fully-automated deep convolutional neural networks (i.e., U-Net) with LiDAR derivatives to identify relict charcoal hearths (RCHs), a type of historical land use feature. Results show that slope, hillshade, and Visualization for Archaeological Topography (VAT) rasters work well in six localized test regions (spatial scale: <1.5 km<sup>2</sup>, best F1 score: 95.5%), but also at broader extents at the town level (spatial scale: 493 km<sup>2</sup>, best F1 score: 86%). The model performed best in areas with deciduous forest and high slope terrain (e.g., >15 degrees) (F1 score: 86.8%) compared to coniferous forest and low slope terrain (e.g., <15 degrees) (F1 score: 70.1%). Overall, our results contribute to current methodological discussions regarding automated extraction of historical cultural features using deep learning and LiDAR.Ji Won SuhEli AndersonWilliam OuimetKatharine M. JohnsonChandi WitharanaMDPI AGarticleremote sensingrelict charcoal hearthsdeep convolutional neural networkssemantic segmentationanthropogenic feature detectionScienceQENRemote Sensing, Vol 13, Iss 4630, p 4630 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	remote sensing relict charcoal hearths deep convolutional neural networks semantic segmentation anthropogenic feature detection Science Q
spellingShingle	remote sensing relict charcoal hearths deep convolutional neural networks semantic segmentation anthropogenic feature detection Science Q Ji Won Suh Eli Anderson William Ouimet Katharine M. Johnson Chandi Witharana Mapping Relict Charcoal Hearths in New England Using Deep Convolutional Neural Networks and LiDAR Data
description	Advanced deep learning methods combined with regional, open access, airborne Light Detection and Ranging (LiDAR) data have great potential to study the spatial extent of historic land use features preserved under the forest canopy throughout New England, a region in the northeastern United States. Mapping anthropogenic features plays a key role in understanding historic land use dynamics during the 17th to early 20th centuries, however previous studies have primarily used manual or semi-automated digitization methods, which are time consuming for broad-scale mapping. This study applies fully-automated deep convolutional neural networks (i.e., U-Net) with LiDAR derivatives to identify relict charcoal hearths (RCHs), a type of historical land use feature. Results show that slope, hillshade, and Visualization for Archaeological Topography (VAT) rasters work well in six localized test regions (spatial scale: <1.5 km<sup>2</sup>, best F1 score: 95.5%), but also at broader extents at the town level (spatial scale: 493 km<sup>2</sup>, best F1 score: 86%). The model performed best in areas with deciduous forest and high slope terrain (e.g., >15 degrees) (F1 score: 86.8%) compared to coniferous forest and low slope terrain (e.g., <15 degrees) (F1 score: 70.1%). Overall, our results contribute to current methodological discussions regarding automated extraction of historical cultural features using deep learning and LiDAR.
format	article
author	Ji Won Suh Eli Anderson William Ouimet Katharine M. Johnson Chandi Witharana
author_facet	Ji Won Suh Eli Anderson William Ouimet Katharine M. Johnson Chandi Witharana
author_sort	Ji Won Suh
title	Mapping Relict Charcoal Hearths in New England Using Deep Convolutional Neural Networks and LiDAR Data
title_short	Mapping Relict Charcoal Hearths in New England Using Deep Convolutional Neural Networks and LiDAR Data
title_full	Mapping Relict Charcoal Hearths in New England Using Deep Convolutional Neural Networks and LiDAR Data
title_fullStr	Mapping Relict Charcoal Hearths in New England Using Deep Convolutional Neural Networks and LiDAR Data
title_full_unstemmed	Mapping Relict Charcoal Hearths in New England Using Deep Convolutional Neural Networks and LiDAR Data
title_sort	mapping relict charcoal hearths in new england using deep convolutional neural networks and lidar data
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/847966d477074eab88e929d3419c6d03
work_keys_str_mv	AT jiwonsuh mappingrelictcharcoalhearthsinnewenglandusingdeepconvolutionalneuralnetworksandlidardata AT elianderson mappingrelictcharcoalhearthsinnewenglandusingdeepconvolutionalneuralnetworksandlidardata AT williamouimet mappingrelictcharcoalhearthsinnewenglandusingdeepconvolutionalneuralnetworksandlidardata AT katharinemjohnson mappingrelictcharcoalhearthsinnewenglandusingdeepconvolutionalneuralnetworksandlidardata AT chandiwitharana mappingrelictcharcoalhearthsinnewenglandusingdeepconvolutionalneuralnetworksandlidardata
_version_	1718410536365850624

Mapping Relict Charcoal Hearths in New England Using Deep Convolutional Neural Networks and LiDAR Data

Ejemplares similares