Deep Learning for Classification of Hyperspectral Data: A Comparative Review

In recent years, deep-learning techniques revolutionized the way remote sensing data are processed. The classification of hyperspectral data is no exception to the rule, but it has intrinsic specificities that make the application of deep learning less straightforward than with other optical data. T...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE geoscience and remote sensing magazine Ročník 7; číslo 2; s. 159 - 173
Hlavní autoři: Audebert, Nicolas, Le Saux, Bertrand, Lefevre, Sebastien
Médium: Journal Article
Jazyk:angličtina
Vydáno: IEEE 01.06.2019
Témata:
Computer Science
Computer Vision and Pattern Recognition
Deep learning
Hyperspectral imaging
Machine learning
Neural and Evolutionary Computing
Sensors
Spatial resolution
ISSN:2473-2397, 2168-6831
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:In recent years, deep-learning techniques revolutionized the way remote sensing data are processed. The classification of hyperspectral data is no exception to the rule, but it has intrinsic specificities that make the application of deep learning less straightforward than with other optical data. This article presents the state of the art of previous machine-learning approaches, reviews the various deeplearning approaches currently proposed for hyperspectral classification, and identifies the problems and difficulties that arise in the implementation of deep neural networks for this task. In particular, the issues of spatial and spectral resolution, data volume, and transfer of models from multimedia images to hyperspectral data are addressed. Additionally, a comparative study of various families of network architectures is provided, and a software toolbox is publicly released to allow experimenting with these methods (https://github.com/nshaud/DeepHyperX). This article is intended for both data scientists with interest in hyperspectral data and remote sensing experts eager to apply deeplearning techniques to their own data set.
ISSN:2473-2397
2168-6831
DOI:10.1109/MGRS.2019.2912563