Uncovering Ecological Patterns with Convolutional Neural Networks

Philip G. Brodrick, Andrew B. Davies, Gregory P. Asner

Research output: Contribution to journalReview article

Abstract

Using remotely sensed imagery to identify biophysical components across landscapes is an important avenue of investigation for ecologists studying ecosystem dynamics. With high-resolution remotely sensed imagery, algorithmic utilization of image context is crucial for accurate identification of biophysical components at large scales. In recent years, convolutional neural networks (CNNs) have become ubiquitous in image processing, and are rapidly becoming more common in ecology. Because the quantity of high-resolution remotely sensed imagery continues to rise, CNNs are increasingly essential tools for large-scale ecosystem analysis. We discuss here the conceptual advantages of CNNs, demonstrate how they can be used by ecologists through distinct examples of their application, and provide a walkthrough of how to use them for ecological applications.

Original languageEnglish (US)
Pages (from-to)734-745
Number of pages12
JournalTrends in Ecology and Evolution
Volume34
Issue number8
DOIs
StatePublished - Aug 2019
Externally publishedYes

Fingerprint

neural networks
imagery
ecologists
ecosystem dynamics
ecosystems
image processing
image analysis
ecology
ecosystem analysis
landscape component

Keywords

  • convolutional neural network
  • deep learning
  • image segmentation
  • machine learning
  • object detection
  • remote sensing

ASJC Scopus subject areas

  • Ecology, Evolution, Behavior and Systematics

Cite this

Uncovering Ecological Patterns with Convolutional Neural Networks. / Brodrick, Philip G.; Davies, Andrew B.; Asner, Gregory P.

In: Trends in Ecology and Evolution, Vol. 34, No. 8, 08.2019, p. 734-745.

Research output: Contribution to journalReview article

Brodrick, Philip G. ; Davies, Andrew B. ; Asner, Gregory P. / Uncovering Ecological Patterns with Convolutional Neural Networks. In: Trends in Ecology and Evolution. 2019 ; Vol. 34, No. 8. pp. 734-745.
@article{e487975b176c4ab5ad7e8f40fb34f2fd,
title = "Uncovering Ecological Patterns with Convolutional Neural Networks",
abstract = "Using remotely sensed imagery to identify biophysical components across landscapes is an important avenue of investigation for ecologists studying ecosystem dynamics. With high-resolution remotely sensed imagery, algorithmic utilization of image context is crucial for accurate identification of biophysical components at large scales. In recent years, convolutional neural networks (CNNs) have become ubiquitous in image processing, and are rapidly becoming more common in ecology. Because the quantity of high-resolution remotely sensed imagery continues to rise, CNNs are increasingly essential tools for large-scale ecosystem analysis. We discuss here the conceptual advantages of CNNs, demonstrate how they can be used by ecologists through distinct examples of their application, and provide a walkthrough of how to use them for ecological applications.",
keywords = "convolutional neural network, deep learning, image segmentation, machine learning, object detection, remote sensing",
author = "Brodrick, {Philip G.} and Davies, {Andrew B.} and Asner, {Gregory P.}",
year = "2019",
month = "8",
doi = "10.1016/j.tree.2019.03.006",
language = "English (US)",
volume = "34",
pages = "734--745",
journal = "Trends in Ecology and Evolution",
issn = "0169-5347",
publisher = "Elsevier Limited",
number = "8",

}

TY - JOUR

T1 - Uncovering Ecological Patterns with Convolutional Neural Networks

AU - Brodrick, Philip G.

AU - Davies, Andrew B.

AU - Asner, Gregory P.

PY - 2019/8

Y1 - 2019/8

N2 - Using remotely sensed imagery to identify biophysical components across landscapes is an important avenue of investigation for ecologists studying ecosystem dynamics. With high-resolution remotely sensed imagery, algorithmic utilization of image context is crucial for accurate identification of biophysical components at large scales. In recent years, convolutional neural networks (CNNs) have become ubiquitous in image processing, and are rapidly becoming more common in ecology. Because the quantity of high-resolution remotely sensed imagery continues to rise, CNNs are increasingly essential tools for large-scale ecosystem analysis. We discuss here the conceptual advantages of CNNs, demonstrate how they can be used by ecologists through distinct examples of their application, and provide a walkthrough of how to use them for ecological applications.

AB - Using remotely sensed imagery to identify biophysical components across landscapes is an important avenue of investigation for ecologists studying ecosystem dynamics. With high-resolution remotely sensed imagery, algorithmic utilization of image context is crucial for accurate identification of biophysical components at large scales. In recent years, convolutional neural networks (CNNs) have become ubiquitous in image processing, and are rapidly becoming more common in ecology. Because the quantity of high-resolution remotely sensed imagery continues to rise, CNNs are increasingly essential tools for large-scale ecosystem analysis. We discuss here the conceptual advantages of CNNs, demonstrate how they can be used by ecologists through distinct examples of their application, and provide a walkthrough of how to use them for ecological applications.

KW - convolutional neural network

KW - deep learning

KW - image segmentation

KW - machine learning

KW - object detection

KW - remote sensing

UR - http://www.scopus.com/inward/record.url?scp=85065237151&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85065237151&partnerID=8YFLogxK

U2 - 10.1016/j.tree.2019.03.006

DO - 10.1016/j.tree.2019.03.006

M3 - Review article

C2 - 31078331

AN - SCOPUS:85065237151

VL - 34

SP - 734

EP - 745

JO - Trends in Ecology and Evolution

JF - Trends in Ecology and Evolution

SN - 0169-5347

IS - 8

ER -