Two-stage compression of hyperspectral images with enhanced classification performance

Chul Hee Lee, Sungwook Youn, Eunjae Lee, Taeuk Jeong, Joan Serra-Sagristà

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Most compression methods for hyperspectral images have been optimized to minimize mean squared errors. However, this kind of compression method may not retain all discriminant information, which is important if hyperspectral images are to be used to distinguish among classes. In this paper, we propose a two-stage compression method for hyperspectral images with encoding residual discriminant information. In the proposed method, we first apply a compression method to hyperspectral images, producing compressed image data. From the compressed image data, we produce reconstructed images. Then we generate residual images by subtracting the reconstructed images from the original images. We also apply a feature extraction method to the original images, which produces a set of feature vectors. By applying these feature vectors to the residual images, we generate discriminant feature images which provide the discriminant information missed by the compression method. In the proposed method, these discriminant feature images are also encoded. Experiments with AVIRIS data show that the proposed method provides better compression efficiency and improved classification accuracy than other compression methods.

Original languageEnglish
Title of host publicationRemotely Sensed Data Compression, Communications, and Processing XII
EditorsChulhee Lee, Bormin Huang, Chein-I Chang
PublisherSPIE
ISBN (Electronic)9781510601154
DOIs
Publication statusPublished - 2016 Jan 1
EventRemotely Sensed Data Compression, Communications, and Processing XII - Baltimore, United States
Duration: 2016 Apr 202016 Apr 21

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume9874
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Other

OtherRemotely Sensed Data Compression, Communications, and Processing XII
CountryUnited States
CityBaltimore
Period16/4/2016/4/21

Fingerprint

Hyperspectral Image
Compression
Discriminant
Feature extraction
Feature Vector
Experiments
Mean Squared Error
Feature Extraction
Encoding
pattern recognition
Minimise

All Science Journal Classification (ASJC) codes

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Cite this

Lee, C. H., Youn, S., Lee, E., Jeong, T., & Serra-Sagristà, J. (2016). Two-stage compression of hyperspectral images with enhanced classification performance. In C. Lee, B. Huang, & C-I. Chang (Eds.), Remotely Sensed Data Compression, Communications, and Processing XII [98740A] (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 9874). SPIE. https://doi.org/10.1117/12.2225568
Lee, Chul Hee ; Youn, Sungwook ; Lee, Eunjae ; Jeong, Taeuk ; Serra-Sagristà, Joan. / Two-stage compression of hyperspectral images with enhanced classification performance. Remotely Sensed Data Compression, Communications, and Processing XII. editor / Chulhee Lee ; Bormin Huang ; Chein-I Chang. SPIE, 2016. (Proceedings of SPIE - The International Society for Optical Engineering).
@inproceedings{e8244cf61dd1485c91e0cc75ada5f95d,
title = "Two-stage compression of hyperspectral images with enhanced classification performance",
abstract = "Most compression methods for hyperspectral images have been optimized to minimize mean squared errors. However, this kind of compression method may not retain all discriminant information, which is important if hyperspectral images are to be used to distinguish among classes. In this paper, we propose a two-stage compression method for hyperspectral images with encoding residual discriminant information. In the proposed method, we first apply a compression method to hyperspectral images, producing compressed image data. From the compressed image data, we produce reconstructed images. Then we generate residual images by subtracting the reconstructed images from the original images. We also apply a feature extraction method to the original images, which produces a set of feature vectors. By applying these feature vectors to the residual images, we generate discriminant feature images which provide the discriminant information missed by the compression method. In the proposed method, these discriminant feature images are also encoded. Experiments with AVIRIS data show that the proposed method provides better compression efficiency and improved classification accuracy than other compression methods.",
author = "Lee, {Chul Hee} and Sungwook Youn and Eunjae Lee and Taeuk Jeong and Joan Serra-Sagrist{\`a}",
year = "2016",
month = "1",
day = "1",
doi = "10.1117/12.2225568",
language = "English",
series = "Proceedings of SPIE - The International Society for Optical Engineering",
publisher = "SPIE",
editor = "Chulhee Lee and Bormin Huang and Chein-I Chang",
booktitle = "Remotely Sensed Data Compression, Communications, and Processing XII",
address = "United States",

}

Lee, CH, Youn, S, Lee, E, Jeong, T & Serra-Sagristà, J 2016, Two-stage compression of hyperspectral images with enhanced classification performance. in C Lee, B Huang & C-I Chang (eds), Remotely Sensed Data Compression, Communications, and Processing XII., 98740A, Proceedings of SPIE - The International Society for Optical Engineering, vol. 9874, SPIE, Remotely Sensed Data Compression, Communications, and Processing XII, Baltimore, United States, 16/4/20. https://doi.org/10.1117/12.2225568

Two-stage compression of hyperspectral images with enhanced classification performance. / Lee, Chul Hee; Youn, Sungwook; Lee, Eunjae; Jeong, Taeuk; Serra-Sagristà, Joan.

Remotely Sensed Data Compression, Communications, and Processing XII. ed. / Chulhee Lee; Bormin Huang; Chein-I Chang. SPIE, 2016. 98740A (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 9874).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Two-stage compression of hyperspectral images with enhanced classification performance

AU - Lee, Chul Hee

AU - Youn, Sungwook

AU - Lee, Eunjae

AU - Jeong, Taeuk

AU - Serra-Sagristà, Joan

PY - 2016/1/1

Y1 - 2016/1/1

N2 - Most compression methods for hyperspectral images have been optimized to minimize mean squared errors. However, this kind of compression method may not retain all discriminant information, which is important if hyperspectral images are to be used to distinguish among classes. In this paper, we propose a two-stage compression method for hyperspectral images with encoding residual discriminant information. In the proposed method, we first apply a compression method to hyperspectral images, producing compressed image data. From the compressed image data, we produce reconstructed images. Then we generate residual images by subtracting the reconstructed images from the original images. We also apply a feature extraction method to the original images, which produces a set of feature vectors. By applying these feature vectors to the residual images, we generate discriminant feature images which provide the discriminant information missed by the compression method. In the proposed method, these discriminant feature images are also encoded. Experiments with AVIRIS data show that the proposed method provides better compression efficiency and improved classification accuracy than other compression methods.

AB - Most compression methods for hyperspectral images have been optimized to minimize mean squared errors. However, this kind of compression method may not retain all discriminant information, which is important if hyperspectral images are to be used to distinguish among classes. In this paper, we propose a two-stage compression method for hyperspectral images with encoding residual discriminant information. In the proposed method, we first apply a compression method to hyperspectral images, producing compressed image data. From the compressed image data, we produce reconstructed images. Then we generate residual images by subtracting the reconstructed images from the original images. We also apply a feature extraction method to the original images, which produces a set of feature vectors. By applying these feature vectors to the residual images, we generate discriminant feature images which provide the discriminant information missed by the compression method. In the proposed method, these discriminant feature images are also encoded. Experiments with AVIRIS data show that the proposed method provides better compression efficiency and improved classification accuracy than other compression methods.

UR - http://www.scopus.com/inward/record.url?scp=84991466991&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84991466991&partnerID=8YFLogxK

U2 - 10.1117/12.2225568

DO - 10.1117/12.2225568

M3 - Conference contribution

AN - SCOPUS:84991466991

T3 - Proceedings of SPIE - The International Society for Optical Engineering

BT - Remotely Sensed Data Compression, Communications, and Processing XII

A2 - Lee, Chulhee

A2 - Huang, Bormin

A2 - Chang, Chein-I

PB - SPIE

ER -

Lee CH, Youn S, Lee E, Jeong T, Serra-Sagristà J. Two-stage compression of hyperspectral images with enhanced classification performance. In Lee C, Huang B, Chang C-I, editors, Remotely Sensed Data Compression, Communications, and Processing XII. SPIE. 2016. 98740A. (Proceedings of SPIE - The International Society for Optical Engineering). https://doi.org/10.1117/12.2225568