By Topic

An Operational Approach to PCA+JPEG2000 Compression of Hyperspectral Imagery

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

3 Author(s)
Qian Du ; Dept. of Electr. & Comput. Eng., Mississippi State Univ., Starkville, MS, USA ; Nam Ly ; Fowler, J.E.

Lossy-compression algorithms typically adopted for hyperspectral remote-sensing imagery-such as JPEG2000-usually produce a monotonically increasing signal-to-noise ratio (SNR) for increasing bitrate. Consequently, it is a common philosophy to employ as large a bitrate as possible so as to obtain the highest achievable SNR. However, it has been observed previously that a higher SNR may not necessarily correspond to better performance at data-analysis tasks, such as classification, anomaly detection, or linear unmixing. Considered specifically is the coupling of JPEG2000 with principal component analysis for spectral decorrelation such that only a few principal components are retained, and, for this compression paradigm, a technique to determine an operational bitrate is proposed with the aim of preserving both the majority of information in a dataset as well as its anomalous pixels. This operational bitrate may be much less than the largest bitrate that the system can allow. Experimental results show that classification and unmixing applied to reconstructed data after compression at this operational bitrate result in performance that is the same as or better than that achieved at higher bitrates; meanwhile, removal and lossless storage of anomalies prior to compression results in their perfect preservation in the reconstructed dataset.

Published in:

Selected Topics in Applied Earth Observations and Remote Sensing, IEEE Journal of  (Volume:7 ,  Issue: 6 )