FuseNet: End-to-end multispectral VHR image fusion and classification

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

1 Downloads (Pure)

Abstract

Classification of very high resolution (VHR) satellite images faces two major challenges: 1) inherent low intra-class and high inter-class spectral similarities and 2) mismatching resolution of available bands. Conventional methods have addressed these challenges by adopting separate stages of image fusion and spatial feature extraction steps. These steps, however, are not jointly optimizing the classification task at hand. We propose a single-stage framework embedding these processing stages in a multiresolution convolutional network. The network, called FuseNet, aims to match the resolution of the panchromatic and multispectral bands in a VHR image using convolutional layers with corresponding downsampling and upsampling operations. We compared FuseNet against the use of separate processing steps for image fusion, such as pansharpening and resampling through interpolation. We also analyzed the sensitivity of the classification performance of FuseNet to a selected number of its hyperparameters. Results show that FuseNet surpasses conventional methods.

Original languageEnglish
Title of host publication2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings
PublisherIEEE
Pages2091-2094
Number of pages4
Volume2018-July
ISBN (Electronic)9781538671504
DOIs
Publication statusPublished - 31 Oct 2018
Event38th IEEE International Geoscience and Remote Sensing Symposium 2018: Observing, Understanding and Forcasting the Dynamics of Our Planet - Feria Valencia Convention & Exhibition Center, Valencia, Spain
Duration: 22 Jul 201827 Jul 2018
Conference number: 38
https://www.igarss2018.org/

Conference

Conference38th IEEE International Geoscience and Remote Sensing Symposium 2018
Abbreviated titleIGARSS 2018
CountrySpain
CityValencia
Period22/07/1827/07/18
Internet address

Fingerprint

Image fusion
Image classification
image resolution
Image resolution
Processing
interpolation
Feature extraction
Interpolation
Satellites
method

Keywords

  • Convolutional networks
  • Deep learning
  • Image fusion
  • Land cover classification
  • VHR image

Cite this

Bergado, J. R., Persello, C., & Stein, A. (2018). FuseNet: End-to-end multispectral VHR image fusion and classification. In 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings (Vol. 2018-July, pp. 2091-2094). [8519214] IEEE. https://doi.org/10.1109/IGARSS.2018.8519214
Bergado, J.R. ; Persello, C. ; Stein, A. / FuseNet : End-to-end multispectral VHR image fusion and classification. 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings. Vol. 2018-July IEEE, 2018. pp. 2091-2094
@inproceedings{f46bd4857bc84c3082d41bb86c0262ce,
title = "FuseNet: End-to-end multispectral VHR image fusion and classification",
abstract = "Classification of very high resolution (VHR) satellite images faces two major challenges: 1) inherent low intra-class and high inter-class spectral similarities and 2) mismatching resolution of available bands. Conventional methods have addressed these challenges by adopting separate stages of image fusion and spatial feature extraction steps. These steps, however, are not jointly optimizing the classification task at hand. We propose a single-stage framework embedding these processing stages in a multiresolution convolutional network. The network, called FuseNet, aims to match the resolution of the panchromatic and multispectral bands in a VHR image using convolutional layers with corresponding downsampling and upsampling operations. We compared FuseNet against the use of separate processing steps for image fusion, such as pansharpening and resampling through interpolation. We also analyzed the sensitivity of the classification performance of FuseNet to a selected number of its hyperparameters. Results show that FuseNet surpasses conventional methods.",
keywords = "Convolutional networks, Deep learning, Image fusion, Land cover classification, VHR image",
author = "J.R. Bergado and C. Persello and A. Stein",
year = "2018",
month = "10",
day = "31",
doi = "10.1109/IGARSS.2018.8519214",
language = "English",
volume = "2018-July",
pages = "2091--2094",
booktitle = "2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings",
publisher = "IEEE",
address = "United States",

}

Bergado, JR, Persello, C & Stein, A 2018, FuseNet: End-to-end multispectral VHR image fusion and classification. in 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings. vol. 2018-July, 8519214, IEEE, pp. 2091-2094, 38th IEEE International Geoscience and Remote Sensing Symposium 2018, Valencia, Spain, 22/07/18. https://doi.org/10.1109/IGARSS.2018.8519214

FuseNet : End-to-end multispectral VHR image fusion and classification. / Bergado, J.R.; Persello, C.; Stein, A.

2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings. Vol. 2018-July IEEE, 2018. p. 2091-2094 8519214.

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - FuseNet

T2 - End-to-end multispectral VHR image fusion and classification

AU - Bergado, J.R.

AU - Persello, C.

AU - Stein, A.

PY - 2018/10/31

Y1 - 2018/10/31

N2 - Classification of very high resolution (VHR) satellite images faces two major challenges: 1) inherent low intra-class and high inter-class spectral similarities and 2) mismatching resolution of available bands. Conventional methods have addressed these challenges by adopting separate stages of image fusion and spatial feature extraction steps. These steps, however, are not jointly optimizing the classification task at hand. We propose a single-stage framework embedding these processing stages in a multiresolution convolutional network. The network, called FuseNet, aims to match the resolution of the panchromatic and multispectral bands in a VHR image using convolutional layers with corresponding downsampling and upsampling operations. We compared FuseNet against the use of separate processing steps for image fusion, such as pansharpening and resampling through interpolation. We also analyzed the sensitivity of the classification performance of FuseNet to a selected number of its hyperparameters. Results show that FuseNet surpasses conventional methods.

AB - Classification of very high resolution (VHR) satellite images faces two major challenges: 1) inherent low intra-class and high inter-class spectral similarities and 2) mismatching resolution of available bands. Conventional methods have addressed these challenges by adopting separate stages of image fusion and spatial feature extraction steps. These steps, however, are not jointly optimizing the classification task at hand. We propose a single-stage framework embedding these processing stages in a multiresolution convolutional network. The network, called FuseNet, aims to match the resolution of the panchromatic and multispectral bands in a VHR image using convolutional layers with corresponding downsampling and upsampling operations. We compared FuseNet against the use of separate processing steps for image fusion, such as pansharpening and resampling through interpolation. We also analyzed the sensitivity of the classification performance of FuseNet to a selected number of its hyperparameters. Results show that FuseNet surpasses conventional methods.

KW - Convolutional networks

KW - Deep learning

KW - Image fusion

KW - Land cover classification

KW - VHR image

UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.1109/IGARSS.2018.8519214

UR - https://ezproxy2.utwente.nl/login?url=https://library.itc.utwente.nl/login/2019/chap/bergado_fus.pdf

U2 - 10.1109/IGARSS.2018.8519214

DO - 10.1109/IGARSS.2018.8519214

M3 - Conference contribution

VL - 2018-July

SP - 2091

EP - 2094

BT - 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings

PB - IEEE

ER -

Bergado JR, Persello C, Stein A. FuseNet: End-to-end multispectral VHR image fusion and classification. In 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings. Vol. 2018-July. IEEE. 2018. p. 2091-2094. 8519214 https://doi.org/10.1109/IGARSS.2018.8519214