Robust joint sparsity model for hyperspectral image classification

Shaoguang Huang, Hongyan Zhang, Wenzhi Liao, Aleksandra Pizurica

Research output: Contribution to conferencePaper

Abstract

Sparsity-based classification methods have been widely used in hyperspectral image (HSI) classification. These methods typically assumed Gaussian noise, neglecting the fact that HSIs are often corrupted by different types of noise in practice. In this paper, we develop a robust super-pixel level joint sparse representation classification model (RSJSRC) to address the mixed noise problem in sparsity-based HSI classification. Our method takes into account both Gaussian and sparse noise. Experimental results on simulated and real data demonstrate the efficiency of the proposed method and clear benefits from the introduced mixed-noise model.

Conference

Conference24th IEEE International Conference on Image Processing (ICIP) 2017
Abbreviated titleICIP2017
CountryChina
CityBeijing
Period17/09/1720/09/17

Fingerprint

Image classification
Pixels

Keywords

  • robust classification
  • hyperspectral image
  • super-pixel segmentation
  • sparse representation
  • gaussian noise
  • sparse matrices
  • optimization
  • image classification
  • geophysical image processing

Cite this

Huang, S., Zhang, H., Liao, W., & Pizurica, A. (2018). Robust joint sparsity model for hyperspectral image classification. 3130-3134. Paper presented at 24th IEEE International Conference on Image Processing (ICIP) 2017, Beijing, China. https://doi.org/10.1109/ICIP.2017.8296859
Huang, Shaoguang ; Zhang, Hongyan ; Liao, Wenzhi ; Pizurica, Aleksandra. / Robust joint sparsity model for hyperspectral image classification. Paper presented at 24th IEEE International Conference on Image Processing (ICIP) 2017, Beijing, China.5 p.
@conference{b8eb3a03d8984831a7918f5c537cabd6,
title = "Robust joint sparsity model for hyperspectral image classification",
abstract = "Sparsity-based classification methods have been widely used in hyperspectral image (HSI) classification. These methods typically assumed Gaussian noise, neglecting the fact that HSIs are often corrupted by different types of noise in practice. In this paper, we develop a robust super-pixel level joint sparse representation classification model (RSJSRC) to address the mixed noise problem in sparsity-based HSI classification. Our method takes into account both Gaussian and sparse noise. Experimental results on simulated and real data demonstrate the efficiency of the proposed method and clear benefits from the introduced mixed-noise model.",
keywords = "robust classification, hyperspectral image, super-pixel segmentation, sparse representation, gaussian noise, sparse matrices, optimization, image classification, geophysical image processing",
author = "Shaoguang Huang and Hongyan Zhang and Wenzhi Liao and Aleksandra Pizurica",
note = "{\circledC} 2018 IEEE.; 24th IEEE International Conference on Image Processing (ICIP) 2017, ICIP2017 ; Conference date: 17-09-2017 Through 20-09-2017",
year = "2018",
month = "2",
day = "22",
doi = "10.1109/ICIP.2017.8296859",
language = "English",
pages = "3130--3134",

}

Huang, S, Zhang, H, Liao, W & Pizurica, A 2018, 'Robust joint sparsity model for hyperspectral image classification' Paper presented at 24th IEEE International Conference on Image Processing (ICIP) 2017, Beijing, China, 17/09/17 - 20/09/17, pp. 3130-3134. https://doi.org/10.1109/ICIP.2017.8296859

Robust joint sparsity model for hyperspectral image classification. / Huang, Shaoguang; Zhang, Hongyan; Liao, Wenzhi; Pizurica, Aleksandra.

2018. 3130-3134 Paper presented at 24th IEEE International Conference on Image Processing (ICIP) 2017, Beijing, China.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Robust joint sparsity model for hyperspectral image classification

AU - Huang, Shaoguang

AU - Zhang, Hongyan

AU - Liao, Wenzhi

AU - Pizurica, Aleksandra

N1 - © 2018 IEEE.

PY - 2018/2/22

Y1 - 2018/2/22

N2 - Sparsity-based classification methods have been widely used in hyperspectral image (HSI) classification. These methods typically assumed Gaussian noise, neglecting the fact that HSIs are often corrupted by different types of noise in practice. In this paper, we develop a robust super-pixel level joint sparse representation classification model (RSJSRC) to address the mixed noise problem in sparsity-based HSI classification. Our method takes into account both Gaussian and sparse noise. Experimental results on simulated and real data demonstrate the efficiency of the proposed method and clear benefits from the introduced mixed-noise model.

AB - Sparsity-based classification methods have been widely used in hyperspectral image (HSI) classification. These methods typically assumed Gaussian noise, neglecting the fact that HSIs are often corrupted by different types of noise in practice. In this paper, we develop a robust super-pixel level joint sparse representation classification model (RSJSRC) to address the mixed noise problem in sparsity-based HSI classification. Our method takes into account both Gaussian and sparse noise. Experimental results on simulated and real data demonstrate the efficiency of the proposed method and clear benefits from the introduced mixed-noise model.

KW - robust classification

KW - hyperspectral image

KW - super-pixel segmentation

KW - sparse representation

KW - gaussian noise

KW - sparse matrices

KW - optimization

KW - image classification

KW - geophysical image processing

UR - http://hdl.handle.net/1854/LU-8529659

U2 - 10.1109/ICIP.2017.8296859

DO - 10.1109/ICIP.2017.8296859

M3 - Paper

SP - 3130

EP - 3134

ER -

Huang S, Zhang H, Liao W, Pizurica A. Robust joint sparsity model for hyperspectral image classification. 2018. Paper presented at 24th IEEE International Conference on Image Processing (ICIP) 2017, Beijing, China. https://doi.org/10.1109/ICIP.2017.8296859