Sparse feature extraction for hyperspectral image classification

Lu Wang, Xiaoming Xie, Wei Li, Qian Du, Guojun Li

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

8 Citations (Scopus)

Abstract

Due to the high dimensionality and redundant spectral information in a hyperspectral image (HSI), principal component analysis (PCA) and linear discriminant analysis (LDA) are commonly-used for its feature extraction. By converting PCA and LDA to regression problems and imposing l1-norm constraint on the regression coefficients, sparse principal component analysis (SPCA) and sparse discriminant analysis (SDA) have been developed for improved feature extraction. Furthermore, recently sparse tensor discriminant analysis (STDA), reserving useful structural information and obtaining multiple interrelated is also proposed. Their performance in HSI classification is investigated in this paper. Experiment results demonstrate the effectiveness of these sparse feature extraction methods, especially for STDA, which outperforms the traditional linear counterparts without maintaining spatial relationships among pixels, such as PCA and LDA.

Original languageEnglish
Title of host publication2015 IEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2015 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1067-1070
Number of pages4
ISBN (Electronic)9781479919482
DOIs
Publication statusPublished - 31 Aug 2015
Externally publishedYes
EventIEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2015 - Chengdu, China
Duration: 12 Jul 201515 Jul 2015

Publication series

Name2015 IEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2015 - Proceedings

Conference

ConferenceIEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2015
Country/TerritoryChina
CityChengdu
Period12/07/1515/07/15

Keywords

  • Sparse projections
  • elastic net
  • feature extraction
  • hyperspectral imagery
  • tensor decomposition

Fingerprint

Dive into the research topics of 'Sparse feature extraction for hyperspectral image classification'. Together they form a unique fingerprint.

Cite this