Fusion of spectral and spatial information for classification of hyperspectral remote-sensed imagery by local graph

Wenzhi Liao, Mauro Dalla Mura, Jocelyn Chanussot, Aleksandra Pizurica

Research output: Contribution to journalArticle

20 Citations (Scopus)

Abstract

Hyperspectral imagery contains a wealth of spectral and spatial information that can improve target detection and recognition performance. Conventional feature extraction methods cannot fully exploit both spectral and spatial information. Data fusion by simply stacking different feature sources together does not take into account the differences between feature sources. In this paper, a local graph-based fusion (LGF) method is proposed to couple dimension reduction and feature fusion of the spectral information (i.e., the spectra in the HS image) and the spatial information [extracted by morphological profiles (MPs)]. In the proposed method, the fusion graph is built on the full data by moving a sliding window from the first pixel to the last one. This yields a clear improvement over a previous approach with fusion graph built on randomly selected samples. Experimental results on real hyperspectral images are very encouraging. Compared to the methods using only single feature and stacking all the features together, the proposed LGF method improves the overall classification accuracy on one of the data sets for more than 20% and 5%, respectively.
LanguageEnglish
Pages583-594
Number of pages12
JournalIEEE Journal of Selected Topics in Earth Observation and Remote Sensing
Volume9
Issue number2
DOIs
Publication statusPublished - 1 Dec 2015

Fingerprint

Data fusion
Target tracking
Feature extraction
imagery
Pixels
stacking
extraction method
sliding
pixel
method

Keywords

  • spatial information
  • SVMS
  • graph
  • data fusion
  • high resolution
  • remote sensing
  • directional morphological profiles
  • attribute profiles
  • feature extraction
  • features
  • filters
  • kernels
  • geophysical image processing

Cite this

@article{fe6ec93c31c54070a7005950fbcf76f0,
title = "Fusion of spectral and spatial information for classification of hyperspectral remote-sensed imagery by local graph",
abstract = "Hyperspectral imagery contains a wealth of spectral and spatial information that can improve target detection and recognition performance. Conventional feature extraction methods cannot fully exploit both spectral and spatial information. Data fusion by simply stacking different feature sources together does not take into account the differences between feature sources. In this paper, a local graph-based fusion (LGF) method is proposed to couple dimension reduction and feature fusion of the spectral information (i.e., the spectra in the HS image) and the spatial information [extracted by morphological profiles (MPs)]. In the proposed method, the fusion graph is built on the full data by moving a sliding window from the first pixel to the last one. This yields a clear improvement over a previous approach with fusion graph built on randomly selected samples. Experimental results on real hyperspectral images are very encouraging. Compared to the methods using only single feature and stacking all the features together, the proposed LGF method improves the overall classification accuracy on one of the data sets for more than 20{\%} and 5{\%}, respectively.",
keywords = "spatial information, SVMS, graph, data fusion, high resolution, remote sensing, directional morphological profiles, attribute profiles, feature extraction, features, filters, kernels, geophysical image processing",
author = "Wenzhi Liao and {Dalla Mura}, Mauro and Jocelyn Chanussot and Aleksandra Pizurica",
year = "2015",
month = "12",
day = "1",
doi = "10.1109/JSTARS.2015.2498664",
language = "English",
volume = "9",
pages = "583--594",
journal = "IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing",
issn = "1939-1404",
publisher = "IEEE",
number = "2",

}

Fusion of spectral and spatial information for classification of hyperspectral remote-sensed imagery by local graph. / Liao, Wenzhi; Dalla Mura, Mauro; Chanussot, Jocelyn; Pizurica, Aleksandra.

In: IEEE Journal of Selected Topics in Earth Observation and Remote Sensing, Vol. 9, No. 2, 01.12.2015, p. 583-594.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Fusion of spectral and spatial information for classification of hyperspectral remote-sensed imagery by local graph

AU - Liao, Wenzhi

AU - Dalla Mura, Mauro

AU - Chanussot, Jocelyn

AU - Pizurica, Aleksandra

PY - 2015/12/1

Y1 - 2015/12/1

N2 - Hyperspectral imagery contains a wealth of spectral and spatial information that can improve target detection and recognition performance. Conventional feature extraction methods cannot fully exploit both spectral and spatial information. Data fusion by simply stacking different feature sources together does not take into account the differences between feature sources. In this paper, a local graph-based fusion (LGF) method is proposed to couple dimension reduction and feature fusion of the spectral information (i.e., the spectra in the HS image) and the spatial information [extracted by morphological profiles (MPs)]. In the proposed method, the fusion graph is built on the full data by moving a sliding window from the first pixel to the last one. This yields a clear improvement over a previous approach with fusion graph built on randomly selected samples. Experimental results on real hyperspectral images are very encouraging. Compared to the methods using only single feature and stacking all the features together, the proposed LGF method improves the overall classification accuracy on one of the data sets for more than 20% and 5%, respectively.

AB - Hyperspectral imagery contains a wealth of spectral and spatial information that can improve target detection and recognition performance. Conventional feature extraction methods cannot fully exploit both spectral and spatial information. Data fusion by simply stacking different feature sources together does not take into account the differences between feature sources. In this paper, a local graph-based fusion (LGF) method is proposed to couple dimension reduction and feature fusion of the spectral information (i.e., the spectra in the HS image) and the spatial information [extracted by morphological profiles (MPs)]. In the proposed method, the fusion graph is built on the full data by moving a sliding window from the first pixel to the last one. This yields a clear improvement over a previous approach with fusion graph built on randomly selected samples. Experimental results on real hyperspectral images are very encouraging. Compared to the methods using only single feature and stacking all the features together, the proposed LGF method improves the overall classification accuracy on one of the data sets for more than 20% and 5%, respectively.

KW - spatial information

KW - SVMS

KW - graph

KW - data fusion

KW - high resolution

KW - remote sensing

KW - directional morphological profiles

KW - attribute profiles

KW - feature extraction

KW - features

KW - filters

KW - kernels

KW - geophysical image processing

UR - http://hdl.handle.net/1854/LU-7011493

U2 - 10.1109/JSTARS.2015.2498664

DO - 10.1109/JSTARS.2015.2498664

M3 - Article

VL - 9

SP - 583

EP - 594

JO - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

T2 - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

JF - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing

SN - 1939-1404

IS - 2

ER -