Hyperspectral band selection using crossover based gravitational search algorithm

Jaime Zabalza, Aizhu Zhang, Ping Ma, Sihan Liu, Genyun Sun, Hui Huang, Zhenjie Wang, Chengyan Lin

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Band selection is an important data dimensionality reduction tool in hyperspectral images (HSIs). To identify the most informative subset band from the hundreds of highly corrected bands in HSIs, a novel hyperspectral band selection method using a crossover based gravitational search algorithm (CGSA) is presented in this paper. In this method, the discriminative capability of each band subset is evaluated by a combined optimization criterion, which is constructed based on the overall classification accuracy and the size of the band subset. As the evolution of the criterion, the subset is updated using the V-shaped transfer function based CGSA. Ultimately, the band subset with the best fitness value is selected. Experiments on two public hyperspectral datasets, i.e. the Indian Pines dataset and the Pavia University dataset, have been conducted to test the performance of the proposed method. Comparing experimental results against the basic GSA and the PSOGSA (hybrid PSO and GSA) revealed that all of the three GSA variants can considerably reduce the band dimensionality of HSIs without damaging their classification accuracy. Moreover, the CGSA shows superiority on both the effectiveness and efficiency compared to the other two GSA variants.
LanguageEnglish
JournalIET Image Processing
Early online date17 Aug 2018
DOIs
Publication statusE-pub ahead of print - 17 Aug 2018

Fingerprint

Set theory
Particle swarm optimization (PSO)
Transfer functions
Data reduction
Experiments

Keywords

  • hyperspectral images
  • image classification
  • particle swarm optimisation

Cite this

Zabalza, Jaime ; Zhang, Aizhu ; Ma, Ping ; Liu, Sihan ; Sun, Genyun ; Huang, Hui ; Wang, Zhenjie ; Lin, Chengyan. / Hyperspectral band selection using crossover based gravitational search algorithm. In: IET Image Processing. 2018.
@article{a175325e502242b487e01553ae82c383,
title = "Hyperspectral band selection using crossover based gravitational search algorithm",
abstract = "Band selection is an important data dimensionality reduction tool in hyperspectral images (HSIs). To identify the most informative subset band from the hundreds of highly corrected bands in HSIs, a novel hyperspectral band selection method using a crossover based gravitational search algorithm (CGSA) is presented in this paper. In this method, the discriminative capability of each band subset is evaluated by a combined optimization criterion, which is constructed based on the overall classification accuracy and the size of the band subset. As the evolution of the criterion, the subset is updated using the V-shaped transfer function based CGSA. Ultimately, the band subset with the best fitness value is selected. Experiments on two public hyperspectral datasets, i.e. the Indian Pines dataset and the Pavia University dataset, have been conducted to test the performance of the proposed method. Comparing experimental results against the basic GSA and the PSOGSA (hybrid PSO and GSA) revealed that all of the three GSA variants can considerably reduce the band dimensionality of HSIs without damaging their classification accuracy. Moreover, the CGSA shows superiority on both the effectiveness and efficiency compared to the other two GSA variants.",
keywords = "hyperspectral images, image classification, particle swarm optimisation",
author = "Jaime Zabalza and Aizhu Zhang and Ping Ma and Sihan Liu and Genyun Sun and Hui Huang and Zhenjie Wang and Chengyan Lin",
year = "2018",
month = "8",
day = "17",
doi = "10.1049/iet-ipr.2018.5362",
language = "English",
journal = "IET Image Processing",
issn = "1751-9659",

}

Hyperspectral band selection using crossover based gravitational search algorithm. / Zabalza, Jaime; Zhang, Aizhu; Ma, Ping; Liu, Sihan; Sun, Genyun; Huang, Hui; Wang, Zhenjie; Lin, Chengyan.

In: IET Image Processing, 17.08.2018.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Hyperspectral band selection using crossover based gravitational search algorithm

AU - Zabalza, Jaime

AU - Zhang, Aizhu

AU - Ma, Ping

AU - Liu, Sihan

AU - Sun, Genyun

AU - Huang, Hui

AU - Wang, Zhenjie

AU - Lin, Chengyan

PY - 2018/8/17

Y1 - 2018/8/17

N2 - Band selection is an important data dimensionality reduction tool in hyperspectral images (HSIs). To identify the most informative subset band from the hundreds of highly corrected bands in HSIs, a novel hyperspectral band selection method using a crossover based gravitational search algorithm (CGSA) is presented in this paper. In this method, the discriminative capability of each band subset is evaluated by a combined optimization criterion, which is constructed based on the overall classification accuracy and the size of the band subset. As the evolution of the criterion, the subset is updated using the V-shaped transfer function based CGSA. Ultimately, the band subset with the best fitness value is selected. Experiments on two public hyperspectral datasets, i.e. the Indian Pines dataset and the Pavia University dataset, have been conducted to test the performance of the proposed method. Comparing experimental results against the basic GSA and the PSOGSA (hybrid PSO and GSA) revealed that all of the three GSA variants can considerably reduce the band dimensionality of HSIs without damaging their classification accuracy. Moreover, the CGSA shows superiority on both the effectiveness and efficiency compared to the other two GSA variants.

AB - Band selection is an important data dimensionality reduction tool in hyperspectral images (HSIs). To identify the most informative subset band from the hundreds of highly corrected bands in HSIs, a novel hyperspectral band selection method using a crossover based gravitational search algorithm (CGSA) is presented in this paper. In this method, the discriminative capability of each band subset is evaluated by a combined optimization criterion, which is constructed based on the overall classification accuracy and the size of the band subset. As the evolution of the criterion, the subset is updated using the V-shaped transfer function based CGSA. Ultimately, the band subset with the best fitness value is selected. Experiments on two public hyperspectral datasets, i.e. the Indian Pines dataset and the Pavia University dataset, have been conducted to test the performance of the proposed method. Comparing experimental results against the basic GSA and the PSOGSA (hybrid PSO and GSA) revealed that all of the three GSA variants can considerably reduce the band dimensionality of HSIs without damaging their classification accuracy. Moreover, the CGSA shows superiority on both the effectiveness and efficiency compared to the other two GSA variants.

KW - hyperspectral images

KW - image classification

KW - particle swarm optimisation

U2 - 10.1049/iet-ipr.2018.5362

DO - 10.1049/iet-ipr.2018.5362

M3 - Article

JO - IET Image Processing

T2 - IET Image Processing

JF - IET Image Processing

SN - 1751-9659

ER -