Further Results in the Use of Independent Components Analysis for Target Detection in Hyperspectral Images

Stefan Robila, Pramod K. Varshney

Research output: Contribution to journalConference articleResearchpeer-review

5 Citations (Scopus)

Abstract

The paper presents a novel algorithm based on Independent Component Analysis (ICA) for the detection of small targets present in hyperspectral images. Compared to previous approaches, the algorithm provides two significant improvements. First, an important speedup is obtained by preprocessing the data through spectral screening. Spectral screening is a technique that measures the similarity between pixel vectors by calculating the angle between them. For a certain threshold α, a set of pixel vectors is selected such that the angle between any two of them is larger than α and the angle between any of the pixel vectors not selected and at least one selected vector is smaller than α. In addition to significantly reducing the size of the data, spectral screening reduces the influence of dominating features. The second improvement is the modification of the Infomax algorithm such that the number of components that are produced is lower than the number of initial observations. This change eliminates the need for feature reduction through PCA, and leads to increased accuracy of the results. Results obtained by applying the new algorithm on data from the hyperspectral digital imagery collection experiment (HYDICE) show that, compared with previous ICA based target detection algorithms developed by the authors, the novel approach has an increased efficiency, at the same time achieving a considerable speedup. The experiments confirm the efficiency of ICA as an attractive tool for hyperspectral data processing.

Original languageEnglish
Pages (from-to)186-195
Number of pages10
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume5094
DOIs
StatePublished - 1 Dec 2003
EventPROCEEDINGS OF SPIE SPIE - The International Society for Optical Engineering: Automatic Target Recognition XIII - Orlando, FL, United States
Duration: 22 Apr 200324 Apr 2003

Fingerprint

Hyperspectral Image
Target Detection
Independent component analysis
Independent Component Analysis
Target tracking
Screening
screening
Pixel
Pixels
pixels
Angle
Speedup
Hyperspectral Data
Number of Components
preprocessing
imagery
Experiment
Preprocessing
Eliminate
Experiments

Keywords

  • Hyperspectral imagery
  • Independent component analysis
  • Mutual information based feature separation
  • Undercomplete representation
  • Unsupervised target detection

Cite this

@article{3e5f8c6413ae4dcabe576f9d7e84776c,
title = "Further Results in the Use of Independent Components Analysis for Target Detection in Hyperspectral Images",
abstract = "The paper presents a novel algorithm based on Independent Component Analysis (ICA) for the detection of small targets present in hyperspectral images. Compared to previous approaches, the algorithm provides two significant improvements. First, an important speedup is obtained by preprocessing the data through spectral screening. Spectral screening is a technique that measures the similarity between pixel vectors by calculating the angle between them. For a certain threshold α, a set of pixel vectors is selected such that the angle between any two of them is larger than α and the angle between any of the pixel vectors not selected and at least one selected vector is smaller than α. In addition to significantly reducing the size of the data, spectral screening reduces the influence of dominating features. The second improvement is the modification of the Infomax algorithm such that the number of components that are produced is lower than the number of initial observations. This change eliminates the need for feature reduction through PCA, and leads to increased accuracy of the results. Results obtained by applying the new algorithm on data from the hyperspectral digital imagery collection experiment (HYDICE) show that, compared with previous ICA based target detection algorithms developed by the authors, the novel approach has an increased efficiency, at the same time achieving a considerable speedup. The experiments confirm the efficiency of ICA as an attractive tool for hyperspectral data processing.",
keywords = "Hyperspectral imagery, Independent component analysis, Mutual information based feature separation, Undercomplete representation, Unsupervised target detection",
author = "Stefan Robila and Varshney, {Pramod K.}",
year = "2003",
month = "12",
day = "1",
doi = "10.1117/12.487452",
language = "English",
volume = "5094",
pages = "186--195",
journal = "Proceedings of SPIE - The International Society for Optical Engineering",
issn = "0277-786X",
publisher = "SPIE",

}

Further Results in the Use of Independent Components Analysis for Target Detection in Hyperspectral Images. / Robila, Stefan; Varshney, Pramod K.

In: Proceedings of SPIE - The International Society for Optical Engineering, Vol. 5094, 01.12.2003, p. 186-195.

Research output: Contribution to journalConference articleResearchpeer-review

TY - JOUR

T1 - Further Results in the Use of Independent Components Analysis for Target Detection in Hyperspectral Images

AU - Robila, Stefan

AU - Varshney, Pramod K.

PY - 2003/12/1

Y1 - 2003/12/1

N2 - The paper presents a novel algorithm based on Independent Component Analysis (ICA) for the detection of small targets present in hyperspectral images. Compared to previous approaches, the algorithm provides two significant improvements. First, an important speedup is obtained by preprocessing the data through spectral screening. Spectral screening is a technique that measures the similarity between pixel vectors by calculating the angle between them. For a certain threshold α, a set of pixel vectors is selected such that the angle between any two of them is larger than α and the angle between any of the pixel vectors not selected and at least one selected vector is smaller than α. In addition to significantly reducing the size of the data, spectral screening reduces the influence of dominating features. The second improvement is the modification of the Infomax algorithm such that the number of components that are produced is lower than the number of initial observations. This change eliminates the need for feature reduction through PCA, and leads to increased accuracy of the results. Results obtained by applying the new algorithm on data from the hyperspectral digital imagery collection experiment (HYDICE) show that, compared with previous ICA based target detection algorithms developed by the authors, the novel approach has an increased efficiency, at the same time achieving a considerable speedup. The experiments confirm the efficiency of ICA as an attractive tool for hyperspectral data processing.

AB - The paper presents a novel algorithm based on Independent Component Analysis (ICA) for the detection of small targets present in hyperspectral images. Compared to previous approaches, the algorithm provides two significant improvements. First, an important speedup is obtained by preprocessing the data through spectral screening. Spectral screening is a technique that measures the similarity between pixel vectors by calculating the angle between them. For a certain threshold α, a set of pixel vectors is selected such that the angle between any two of them is larger than α and the angle between any of the pixel vectors not selected and at least one selected vector is smaller than α. In addition to significantly reducing the size of the data, spectral screening reduces the influence of dominating features. The second improvement is the modification of the Infomax algorithm such that the number of components that are produced is lower than the number of initial observations. This change eliminates the need for feature reduction through PCA, and leads to increased accuracy of the results. Results obtained by applying the new algorithm on data from the hyperspectral digital imagery collection experiment (HYDICE) show that, compared with previous ICA based target detection algorithms developed by the authors, the novel approach has an increased efficiency, at the same time achieving a considerable speedup. The experiments confirm the efficiency of ICA as an attractive tool for hyperspectral data processing.

KW - Hyperspectral imagery

KW - Independent component analysis

KW - Mutual information based feature separation

KW - Undercomplete representation

KW - Unsupervised target detection

UR - http://www.scopus.com/inward/record.url?scp=0344944857&partnerID=8YFLogxK

U2 - 10.1117/12.487452

DO - 10.1117/12.487452

M3 - Conference article

VL - 5094

SP - 186

EP - 195

JO - Proceedings of SPIE - The International Society for Optical Engineering

JF - Proceedings of SPIE - The International Society for Optical Engineering

SN - 0277-786X

ER -