Abstract
Since hyperspectral images are very large, it is desirable to compress them before transmission. After receiving the compressed image, decompression is applied before performing image classification and other operations. In this paper, a new processing scheme is proposed, where image transform and quantization are applied for image compression at the transmitter and classification is performed directly on the compressed data at the receiver. The advantage of this scheme is that fewer computations are needed. Computer simulations are performed on hyperspectral imagery.
Original language | English (US) |
---|---|
Title of host publication | WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing |
DOIs | |
State | Published - 2009 |
Externally published | Yes |
Event | WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing - Grenoble, France Duration: Aug 26 2009 → Aug 28 2009 |
Other
Other | WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing |
---|---|
Country | France |
City | Grenoble |
Period | 8/26/09 → 8/28/09 |
Fingerprint
Keywords
- Hyperspectral imagery
- Image classification
- Lossy data compression
- Neural network
ASJC Scopus subject areas
- Computer Networks and Communications
- Computer Vision and Pattern Recognition
- Signal Processing
Cite this
A novel scheme for the compression and classification of hyperspectral images. / Xie, Bei; Bose, Tamal; Merényi, Erzsébet.
WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing. 2009. 5289075.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - A novel scheme for the compression and classification of hyperspectral images
AU - Xie, Bei
AU - Bose, Tamal
AU - Merényi, Erzsébet
PY - 2009
Y1 - 2009
N2 - Since hyperspectral images are very large, it is desirable to compress them before transmission. After receiving the compressed image, decompression is applied before performing image classification and other operations. In this paper, a new processing scheme is proposed, where image transform and quantization are applied for image compression at the transmitter and classification is performed directly on the compressed data at the receiver. The advantage of this scheme is that fewer computations are needed. Computer simulations are performed on hyperspectral imagery.
AB - Since hyperspectral images are very large, it is desirable to compress them before transmission. After receiving the compressed image, decompression is applied before performing image classification and other operations. In this paper, a new processing scheme is proposed, where image transform and quantization are applied for image compression at the transmitter and classification is performed directly on the compressed data at the receiver. The advantage of this scheme is that fewer computations are needed. Computer simulations are performed on hyperspectral imagery.
KW - Hyperspectral imagery
KW - Image classification
KW - Lossy data compression
KW - Neural network
UR - http://www.scopus.com/inward/record.url?scp=72049130299&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=72049130299&partnerID=8YFLogxK
U2 - 10.1109/WHISPERS.2009.5289075
DO - 10.1109/WHISPERS.2009.5289075
M3 - Conference contribution
AN - SCOPUS:72049130299
SN - 9781424446872
BT - WHISPERS '09 - 1st Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing
ER -