Bit allocation for 2D compression of hyperspectral images for classification

Sangwook Lee, Jonghwa Lee, Chul Hee Lee

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

1 Citation (Scopus)

Abstract

In this paper, we propose a bit allocation method for 2D compression of hyperspectral images to enhance classification performance. First, we select a number of classes from original hyperspectral images. It is noted that the classes can be automatically selected by applying an unsupervised segmentation method. Then, we apply a feature extraction method and determine discriminately dominant feature vectors. By examining the feature vectors, we determine the discriminant usefulness of each spectral band. Finally, based on the discriminant usefulness of the spectral bands, we determine bit allocation of each spectral band. Experimental results show that it is possible to enhance the discriminant information at the expense of PSNR. Depending on applications, one can either minimize the mean squared error or choose to preserve the classification capability of the hyperspectral images.

Original languageEnglish
Title of host publicationSatellite Data Compression, Communication, and Processing V
DOIs
Publication statusPublished - 2009 Oct 26
EventSatellite Data Compression, Communication, and Processing V - San Diego, CA, United States
Duration: 2009 Aug 42009 Aug 5

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume7455
ISSN (Print)0277-786X

Other

OtherSatellite Data Compression, Communication, and Processing V
CountryUnited States
CitySan Diego, CA
Period09/8/409/8/5

Fingerprint

Bit Allocation
Hyperspectral Image
spectral bands
Discriminant
Compression
Feature Vector
Feature extraction
Mean Squared Error
pattern recognition
Feature Extraction
Segmentation
Choose
Minimise
Experimental Results
Class

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, S., Lee, J., & Lee, C. H. (2009). Bit allocation for 2D compression of hyperspectral images for classification. In Satellite Data Compression, Communication, and Processing V [745507] (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 7455). https://doi.org/10.1117/12.826958
Lee, Sangwook ; Lee, Jonghwa ; Lee, Chul Hee. / Bit allocation for 2D compression of hyperspectral images for classification. Satellite Data Compression, Communication, and Processing V. 2009. (Proceedings of SPIE - The International Society for Optical Engineering).
@inproceedings{723692ece79b42c59a96f51d7d7e1c5b,
title = "Bit allocation for 2D compression of hyperspectral images for classification",
abstract = "In this paper, we propose a bit allocation method for 2D compression of hyperspectral images to enhance classification performance. First, we select a number of classes from original hyperspectral images. It is noted that the classes can be automatically selected by applying an unsupervised segmentation method. Then, we apply a feature extraction method and determine discriminately dominant feature vectors. By examining the feature vectors, we determine the discriminant usefulness of each spectral band. Finally, based on the discriminant usefulness of the spectral bands, we determine bit allocation of each spectral band. Experimental results show that it is possible to enhance the discriminant information at the expense of PSNR. Depending on applications, one can either minimize the mean squared error or choose to preserve the classification capability of the hyperspectral images.",
author = "Sangwook Lee and Jonghwa Lee and Lee, {Chul Hee}",
year = "2009",
month = "10",
day = "26",
doi = "10.1117/12.826958",
language = "English",
isbn = "9780819477453",
series = "Proceedings of SPIE - The International Society for Optical Engineering",
booktitle = "Satellite Data Compression, Communication, and Processing V",

}

Lee, S, Lee, J & Lee, CH 2009, Bit allocation for 2D compression of hyperspectral images for classification. in Satellite Data Compression, Communication, and Processing V., 745507, Proceedings of SPIE - The International Society for Optical Engineering, vol. 7455, Satellite Data Compression, Communication, and Processing V, San Diego, CA, United States, 09/8/4. https://doi.org/10.1117/12.826958

Bit allocation for 2D compression of hyperspectral images for classification. / Lee, Sangwook; Lee, Jonghwa; Lee, Chul Hee.

Satellite Data Compression, Communication, and Processing V. 2009. 745507 (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 7455).

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

TY - GEN

T1 - Bit allocation for 2D compression of hyperspectral images for classification

AU - Lee, Sangwook

AU - Lee, Jonghwa

AU - Lee, Chul Hee

PY - 2009/10/26

Y1 - 2009/10/26

N2 - In this paper, we propose a bit allocation method for 2D compression of hyperspectral images to enhance classification performance. First, we select a number of classes from original hyperspectral images. It is noted that the classes can be automatically selected by applying an unsupervised segmentation method. Then, we apply a feature extraction method and determine discriminately dominant feature vectors. By examining the feature vectors, we determine the discriminant usefulness of each spectral band. Finally, based on the discriminant usefulness of the spectral bands, we determine bit allocation of each spectral band. Experimental results show that it is possible to enhance the discriminant information at the expense of PSNR. Depending on applications, one can either minimize the mean squared error or choose to preserve the classification capability of the hyperspectral images.

AB - In this paper, we propose a bit allocation method for 2D compression of hyperspectral images to enhance classification performance. First, we select a number of classes from original hyperspectral images. It is noted that the classes can be automatically selected by applying an unsupervised segmentation method. Then, we apply a feature extraction method and determine discriminately dominant feature vectors. By examining the feature vectors, we determine the discriminant usefulness of each spectral band. Finally, based on the discriminant usefulness of the spectral bands, we determine bit allocation of each spectral band. Experimental results show that it is possible to enhance the discriminant information at the expense of PSNR. Depending on applications, one can either minimize the mean squared error or choose to preserve the classification capability of the hyperspectral images.

UR - http://www.scopus.com/inward/record.url?scp=70350150169&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70350150169&partnerID=8YFLogxK

U2 - 10.1117/12.826958

DO - 10.1117/12.826958

M3 - Conference contribution

SN - 9780819477453

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

BT - Satellite Data Compression, Communication, and Processing V

ER -

Lee S, Lee J, Lee CH. Bit allocation for 2D compression of hyperspectral images for classification. In Satellite Data Compression, Communication, and Processing V. 2009. 745507. (Proceedings of SPIE - The International Society for Optical Engineering). https://doi.org/10.1117/12.826958