Skip to main navigation Skip to search Skip to main content

Spectral-spatial classification of hyperspectral image based on locality preserving discriminant analysis

  • Min Han*
  • , Chengkun Zhang
  • , Jun Wang
  • *Corresponding author for this work

Research output: Chapters, Conference Papers, Creative and Literary WorksRGC 32 - Refereed conference paper (with host publication)peer-review

Abstract

In this paper, a spectral-spatial classification method for hyperspectral image based on spatial filtering and feature extraction is proposed. To extract the spatial information that contain spatially homogeneous property and distinct boundary, the original hyperspectral image is processed by an improved bilateral filter firstly. And then the proposed feature extraction algorithm called locality preserving discriminant analysis, which can explore the manifold structure and intrinsic characteristics of the hyperspectral dataset, is used to reduce the dimensionality of both the spectral and spatial features. Finally, a support vector machine with a composite kernel is used to examine the performance of the proposed methods. Experiments results on a hyperspectral dataset demonstrate the effectiveness of the proposed algorithm in the classification tasks.
Original languageEnglish
Title of host publicationAdvances in Neural Networks
Subtitle of host publication13th International Symposium on Neural Networks, ISNN 2016, Proceedings
EditorsLong Cheng, Qingshan Liu, Andrey Ronzhin
PublisherSpringer Verlag
Pages21-29
Volume9719
ISBN (Print)9783319406626
DOIs
Publication statusPublished - 2016
Event13th International Symposium on Neural Networks, ISNN 2016 - St. Petersburg, Russian Federation
Duration: 6 Jul 20168 Jul 2016

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9719
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference13th International Symposium on Neural Networks, ISNN 2016
PlaceRussian Federation
CitySt. Petersburg
Period6/07/168/07/16

Research Keywords

  • Feature extraction
  • Hyperspectral
  • Manifold structure
  • Spatial filtering
  • Support vector machine with a composite kernel

Fingerprint

Dive into the research topics of 'Spectral-spatial classification of hyperspectral image based on locality preserving discriminant analysis'. Together they form a unique fingerprint.

Cite this