Class-based compression algorithm for hyperspectral images

Yang Xinfeng; Hu Xunuo; Nian Yongjian

doi:10.3788/IRLA201645.0228003

The huge amount of hyperspectral images creates challenges for data storage and transmission, thus it is necessary to employ efficient algorithm for hyperspectral images compression. An efficient lossy compression algorithm based on spectral classification was presented in this paper. The C-means algorithm was performed on the hyperspectral images to realize the unsupervised classification. According to the classification map, an adaptive Karhunen-Love transform was performed on each class vector with the same spatial location in the spectral orientation to remove the spectral correlation, and then two dimensional wavelet transform was performed on each principle component. In order to achieve the best rate-distortion performance, the embedded block coding with optimized truncation coding was performed on all the principle components to produce the final bit-stream. Experimental results show that the proposed algorithm outperforms other state-of-the-art algorithms.

Class-based compression algorithm for hyperspectral images

1. School of Computer & Information Engineering,Nanyang Institute of Technology,Nanyang 473000,China;

2. Department of Health Management,Nanyang Medical College,Nanyang 473000,China;

3. School of Biomedical Engineering,Third Military Medical University,Chongqing 400038,China

Received Date: 2015-06-06

Rev Recd Date: 2015-07-10

Publish Date: 2016-02-25

Key words:

Abstract: The huge amount of hyperspectral images creates challenges for data storage and transmission, thus it is necessary to employ efficient algorithm for hyperspectral images compression. An efficient lossy compression algorithm based on spectral classification was presented in this paper. The C-means algorithm was performed on the hyperspectral images to realize the unsupervised classification. According to the classification map, an adaptive Karhunen-Love transform was performed on each class vector with the same spatial location in the spectral orientation to remove the spectral correlation, and then two dimensional wavelet transform was performed on each principle component. In order to achieve the best rate-distortion performance, the embedded block coding with optimized truncation coding was performed on all the principle components to produce the final bit-stream. Experimental results show that the proposed algorithm outperforms other state-of-the-art algorithms.

HTML

Reference (13)

[1]
[2]	Lee H S, Younan N H, King R L. Hyperspectral image cube compression combining JPEG-2000 and spectral decorrelation[C]//Proceedings of the International Geoscience and Remote Sensing Symposium, Toronto, Canada, 2002, 6:3317-3319.
[3]
[4]	Penna B, Tillo T, Magli E, et al. Transform coding techniques for lossy hyperspectral data compression[J]. IEEE Transactions on Geoscience and Remote Sensing, 2007, 45(5):1408-1421.
[5]	Wang L, Wu J J, Jiao L C, et al. Lossy-to-lossless hyperspectral image compression based on multiplierless reversible integer TDLT/KLT[J]. IEEE Geoscience and Remote Sensing Letters, 2009, 6(3):587-591.
[6]
[7]	Fang L J, Nian Y J, Wang Y C. Hyperspectral images compression based on classified KLT[J]. Computer Technology and Development, 2013, 23(11):82-85.
[8]
[9]
[10]	Cagnazzo M, Poggi G, Verdoliva L. Region-based transform coding of multispectral images[J]. IEEE Transactions on Image Processing, 2007, 16(12):2916-2926.
[11]
[12]	Taubman D. High performance scalable image compression with EBCOT[J]. IEEE Transactions on Image Processing, 2000, 9(7):1158-1170.
[13]	Tang X L, Pearlman W A. Three-Dimensional Wavelet-Based Compression of Hyperspectral Images[M]. Berlin:Springer, 2006:273-308.

Class-based compression algorithm for hyperspectral images

doi: 10.3788/IRLA201645.0228003

Abstract

References

Proportional views

通讯作者: 陈斌, bchen63@163.com

Article Metrics

Related

Proportional views