Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Bilateral Kernel Extraction from PCA for Classification of Hyperspectral Images

Bilateral Kernel Extraction from PCA for Classification of Hyperspectral Images The improved spatial and spectral resolution in the advanced Hyperspectral (HS) sensors results in images with rich information per pixel. Hence, the development of efficient spatial–spectral feature extraction (FE) techniques is very crucial for a proper characterization of the objects on ground. In this paper, an attempt has been made to develop a simple, yet effective spatial–spectral FE algorithm. In the proposed approach, the following steps are performed. First, Principal Component Analysis (PCA) was applied on the original hyperspectral image (HSI) and the most significant principal component was extracted. Then, the Bilateral Filter (BF), which acts as an edge-preserving filter, was applied on the selected principal component to extract kernel for each pixel in HSI. The extracted kernel bank is then applied on the original HSI. As in general, the principal component image is edge informative, and the BF is an edge-preserving filter; therefore, the extracted kernel bank can be applied on the original HSI to extract spatial–spectral features. Finally, with the help of these features, the performance of Support Vector Machine (SVM) classifier is evaluated. The proposed approach is validated on three popular hyperspectral data sets, namely, Indian Pines, Pavia University, and Botswana. The experimental results reveal that learning the edge information from a reference image (in the present context PCA) is quite essential, rather than applying the edge-preserving filters directly on the HSI. Theoretically, this holds true, as a unique edge (ground) information exists for an HSI, while in reality, the edges have variations due to variation in reflectance over bands. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png PFG – Journal of Photogrammetry, Remote Sensing and Geoinformation Science Springer Journals

Bilateral Kernel Extraction from PCA for Classification of Hyperspectral Images

Download PDF
17 pages

Loading next page...
 
/lp/springer-journals/bilateral-kernel-extraction-from-pca-for-classification-of-4kG6WVjAmS
Publisher
Springer Journals
Copyright
Copyright © 2019 by Deutsche Gesellschaft für Photogrammetrie, Fernerkundung und Geoinformation (DGPF) e.V.
Subject
Geography; Remote Sensing/Photogrammetry; Geographical Information Systems/Cartography; Signal,Image and Speech Processing; Computer Imaging, Vision, Pattern Recognition and Graphics; Astronomy, Observations and Techniques; Aerospace Technology and Astronautics
ISSN
2512-2789
eISSN
2512-2819
DOI
10.1007/s41064-019-00078-9
Publisher site
See Article on Publisher Site

Abstract

The improved spatial and spectral resolution in the advanced Hyperspectral (HS) sensors results in images with rich information per pixel. Hence, the development of efficient spatial–spectral feature extraction (FE) techniques is very crucial for a proper characterization of the objects on ground. In this paper, an attempt has been made to develop a simple, yet effective spatial–spectral FE algorithm. In the proposed approach, the following steps are performed. First, Principal Component Analysis (PCA) was applied on the original hyperspectral image (HSI) and the most significant principal component was extracted. Then, the Bilateral Filter (BF), which acts as an edge-preserving filter, was applied on the selected principal component to extract kernel for each pixel in HSI. The extracted kernel bank is then applied on the original HSI. As in general, the principal component image is edge informative, and the BF is an edge-preserving filter; therefore, the extracted kernel bank can be applied on the original HSI to extract spatial–spectral features. Finally, with the help of these features, the performance of Support Vector Machine (SVM) classifier is evaluated. The proposed approach is validated on three popular hyperspectral data sets, namely, Indian Pines, Pavia University, and Botswana. The experimental results reveal that learning the edge information from a reference image (in the present context PCA) is quite essential, rather than applying the edge-preserving filters directly on the HSI. Theoretically, this holds true, as a unique edge (ground) information exists for an HSI, while in reality, the edges have variations due to variation in reflectance over bands.

Journal

PFG – Journal of Photogrammetry, Remote Sensing and Geoinformation Science – Springer Journals

Published: Sep 6, 2019

References

  • Classification of hyperspectral datafrom urban areas based on extended morphological profiles
    Benediktsson, JA; Palmason, JA; Sveinsson, JR
  • LIBSVM
    Chang, Chih-Chung; Lin, Chih-Jen
  • Morphological attributeprofiles for the analysis of very high resolution images
    Dalla Mura, M; Benediktsson, JA; Waske, B; Bruzzone, L
  • Classification ofhyperspectral images by using extended morphological attribute profiles and independentcomponent analysis
    Dalla Mura, M; Villa, A; Benediktsson, JA; Chanussot, J; Bruzzone, L
  • Advances in Principal Component Analysis
    Datta, Aloke; Ghosh, Susmita; Ghosh, Ashish
  • A survey on spectral-spatial classifica-tion techniques based on attribute profiles
    Ghamisi, P; Dalla Mura, M; Benediktsson, JA
  • Extinction profiles for the classification of remote sensing data
    Ghamisi, P; Souza, R; Benediktsson, JA; Zhu, XX; Rittner, L; Lotufo, RA
  • Hyperspectral data classification using extended extinction profiles
    Ghamisi, P; Souza, R; Benediktsson, JA; Rittner, L; Lotufo, R; Zhu, XX
  • Three-dimensional wavelet texture feature extraction andclassification for multi/hyperspectral imagery
    Guo, X; Huang, X; Zhang, L
  • A multichannel gray level co-occurrence matrix formulti/hyperspectral image texture representation
    Huang, X; Liu, X; Zhang, L
  • Gabor feature-based collaborative representation for hyperspectralimagery classification
    Jia, S; Shen, L; Li, Q
  • Spectral-spatial hyperspectral image classificationwith edge-preserving filtering
    Kang, X; Li, S; Benediktsson, JA
  • Feature extraction of hyperspectral images with imagefusion and recursive filtering
    Kang, X; Li, S; Benediktsson, JA
  • PCA-based edge-preserving features forhyperspectral image classification
    Kang, X; Xiang, X; Li, S; Benediktsson, JA
  • Visualization of hyperspectral images using bilateral filtering
    Kotwal, K; Chaudhuri, S
  • Hyperspectral image data analysis
    Landgrebe, D
  • Hyperspectral image segmentation using a newbayesian approach with active learning
    Li, J; Bioucas-Dias, JM; Plaza, A
  • Generalized com-posite kernel framework for hyperspectral image classification
    Li, J; Marpu, PR; Plaza, A; Bioucas-Dias, JM; Benediktsson, JA
  • Local binary patterns and extreme learning machine forhyperspectral imagery classification
    Li, W; Chen, C; Su, H; Du, Q
  • Minimum noise fraction versus principalcomponent analysis as a preprocessing step for hyperspectral imagery denoising
    Luo, G; Chen, G; Tian, L; Qin, K; Qian, SE
  • Auto-matic generation of standard deviation attribute profiles for spectral-spatial classificationof remote sensing data
    Marpu, PR; Pedergnana, M; Dalla Mura, M; Benediktsson, JA; Bruzzone, L
  • Classification of hyperspectral remote sensing images with sup-port vector machines
    Melgani, F; Bruzzone, L
  • Hyperspectral image classification with improvedlocal-region filters
    Ran, Q; Li, W; Du, Q; Xiong, M
  • Hyperspectral image preprocessingwith bilateral filter for improving the classification accuracy of support vector machines
    Sahadevan, AS; Routray, A; Das, BS; Ahmad, S
  • Independent component analysis-based dimensionality reductionwith applications in hyperspectral image analysis
    Wang, J; Chang, CI
  • Spectral-spatial classification of hyperspectral images usingjoint bilateral filter and graph cut based model
    Wang, Y; Song, H; Zhang, Y