Deep Fully Convolutional Networks for the Detection of Informal Settlements in VHR Images

Research output: Contribution to journalArticleAcademicpeer-review

32 Citations (Scopus)
5 Downloads (Pure)

Abstract

This letter investigates fully convolutional networks (FCNs) for the detection of informal settlements in very high resolution (VHR) satellite images. Informal settlements or slums are proliferating in developing countries and their detection and classification provides vital information for decision making and planning urban upgrading processes. Distinguishing different urban structures in VHR images is challenging because of the abstract semantic definition of the classes as opposed to the separation of standard land-cover classes. This task requires extraction of texture and spatial features. To this aim, we introduce deep FCNs to perform pixel-wise image labeling by automatically learning a higher level representation of the data. Deep FCNs can learn a hierarchy of features associated to increasing levels of abstraction, from raw pixel values to edges and corners up to complex spatial patterns. We present a deep FCN using dilated convolutions of increasing spatial support. It is capable of learning informative features capturing long-range pixel dependencies while keeping a limited number of network parameters. Experiments carried out on a Quickbird image acquired over the city of Dar es Salaam, Tanzania, show that the proposed FCN outperforms state-of-the-art convolutional networks. Moreover, the computational cost of the proposed technique is significantly lower than standard patch-based architectures.
Original languageEnglish
Pages (from-to)2325-2329
JournalIEEE geoscience and remote sensing letters
Volume14
Issue number12
DOIs
Publication statusPublished - 1 Jan 2017

Fingerprint

informal settlement
image resolution
Image resolution
pixel
Pixels
learning
Urban planning
QuickBird
Convolution
urban planning
Developing countries
Labeling
land cover
Textures
texture
Decision making
developing world
Semantics
decision making
Satellites

Keywords

  • ITC-ISI-JOURNAL-ARTICLE

Cite this

@article{06ba0147b9c94b5a90ea2a250d840e10,
title = "Deep Fully Convolutional Networks for the Detection of Informal Settlements in VHR Images",
abstract = "This letter investigates fully convolutional networks (FCNs) for the detection of informal settlements in very high resolution (VHR) satellite images. Informal settlements or slums are proliferating in developing countries and their detection and classification provides vital information for decision making and planning urban upgrading processes. Distinguishing different urban structures in VHR images is challenging because of the abstract semantic definition of the classes as opposed to the separation of standard land-cover classes. This task requires extraction of texture and spatial features. To this aim, we introduce deep FCNs to perform pixel-wise image labeling by automatically learning a higher level representation of the data. Deep FCNs can learn a hierarchy of features associated to increasing levels of abstraction, from raw pixel values to edges and corners up to complex spatial patterns. We present a deep FCN using dilated convolutions of increasing spatial support. It is capable of learning informative features capturing long-range pixel dependencies while keeping a limited number of network parameters. Experiments carried out on a Quickbird image acquired over the city of Dar es Salaam, Tanzania, show that the proposed FCN outperforms state-of-the-art convolutional networks. Moreover, the computational cost of the proposed technique is significantly lower than standard patch-based architectures.",
keywords = "ITC-ISI-JOURNAL-ARTICLE",
author = "Claudio Persello and Alfred Stein",
year = "2017",
month = "1",
day = "1",
doi = "10.1109/LGRS.2017.2763738",
language = "English",
volume = "14",
pages = "2325--2329",
journal = "IEEE geoscience and remote sensing letters",
issn = "1545-598X",
publisher = "IEEE",
number = "12",

}

Deep Fully Convolutional Networks for the Detection of Informal Settlements in VHR Images. / Persello, Claudio; Stein, Alfred.

In: IEEE geoscience and remote sensing letters, Vol. 14, No. 12, 01.01.2017, p. 2325-2329.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Deep Fully Convolutional Networks for the Detection of Informal Settlements in VHR Images

AU - Persello, Claudio

AU - Stein, Alfred

PY - 2017/1/1

Y1 - 2017/1/1

N2 - This letter investigates fully convolutional networks (FCNs) for the detection of informal settlements in very high resolution (VHR) satellite images. Informal settlements or slums are proliferating in developing countries and their detection and classification provides vital information for decision making and planning urban upgrading processes. Distinguishing different urban structures in VHR images is challenging because of the abstract semantic definition of the classes as opposed to the separation of standard land-cover classes. This task requires extraction of texture and spatial features. To this aim, we introduce deep FCNs to perform pixel-wise image labeling by automatically learning a higher level representation of the data. Deep FCNs can learn a hierarchy of features associated to increasing levels of abstraction, from raw pixel values to edges and corners up to complex spatial patterns. We present a deep FCN using dilated convolutions of increasing spatial support. It is capable of learning informative features capturing long-range pixel dependencies while keeping a limited number of network parameters. Experiments carried out on a Quickbird image acquired over the city of Dar es Salaam, Tanzania, show that the proposed FCN outperforms state-of-the-art convolutional networks. Moreover, the computational cost of the proposed technique is significantly lower than standard patch-based architectures.

AB - This letter investigates fully convolutional networks (FCNs) for the detection of informal settlements in very high resolution (VHR) satellite images. Informal settlements or slums are proliferating in developing countries and their detection and classification provides vital information for decision making and planning urban upgrading processes. Distinguishing different urban structures in VHR images is challenging because of the abstract semantic definition of the classes as opposed to the separation of standard land-cover classes. This task requires extraction of texture and spatial features. To this aim, we introduce deep FCNs to perform pixel-wise image labeling by automatically learning a higher level representation of the data. Deep FCNs can learn a hierarchy of features associated to increasing levels of abstraction, from raw pixel values to edges and corners up to complex spatial patterns. We present a deep FCN using dilated convolutions of increasing spatial support. It is capable of learning informative features capturing long-range pixel dependencies while keeping a limited number of network parameters. Experiments carried out on a Quickbird image acquired over the city of Dar es Salaam, Tanzania, show that the proposed FCN outperforms state-of-the-art convolutional networks. Moreover, the computational cost of the proposed technique is significantly lower than standard patch-based architectures.

KW - ITC-ISI-JOURNAL-ARTICLE

UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.1109/LGRS.2017.2763738

UR - https://ezproxy2.utwente.nl/login?url=https://webapps.itc.utwente.nl/library/2017/isi/persello_deep.pdf

U2 - 10.1109/LGRS.2017.2763738

DO - 10.1109/LGRS.2017.2763738

M3 - Article

VL - 14

SP - 2325

EP - 2329

JO - IEEE geoscience and remote sensing letters

JF - IEEE geoscience and remote sensing letters

SN - 1545-598X

IS - 12

ER -