442
June 2015
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
dynamic dimensionality allocation, and progressive band
selection). However, I felt these chapters are out of place and
should actually have been placed much earlier, probably at
end of Part I.
The author clearly defines the difference between the
hyperspectral image processing and hyperspectral signal
processing as follows:
“
The only difference between hyperspectral image
processing and hyperspectral signal processing is that
the former takes advantage of statistics resulting from
spectral correlation among pixel vectors in an image
cube, while the later processes a hyperspectral signal as
an individual ID signal such as signatures from spectral
libraries or databases without accounting for spectral
correlation among sample signals
”.
And treats them separately as exact methods and approaches
used in hyperspectral- image vs. signal processing throughout
the book this needs to be appreciated by the readers.
A powerful and simple approach to hyperspectral signal
processing is through a number of binary coding methods as
discussed in Chapter 24 such as spectral analysis manager
(SPAM), median partition (MP), halfway partition (HP),
and equal probability partition (EPP). Then, there is vector
signature coding techniques like spectral derivative feature
coding (SDFC) and spectral feature probabilistic coding
(SFDC) (Chapter 25) for signature coding. Finally, some new
techniques of signature coding called progressive signature
coding (PSC) are introduced in Chapter 26.
Second part of the hyperspectral signal processing is
provided in section VII. hyperspectral signal characterization
is defined as “
1D continuous signal processing
” as opposed to
hyperspectral signal coding as “
1D discrete signal processing
”.
Several unique algorithms are discussed for hyperspectral
signal characterization like variable-number variable-
band selection (VNVBS) to analyze a single hyperspectral
signature. This is followed by the widely used Kalman-Filter
based spectral characterization signal processing (KFSCSP)
in Chapter 28, and wavelet-based signature characterization
algorithm (WSCA) in Chapter 29.
The book concludes with applications of hyperspectral
data, such as applications of target detection (Chapter 30),
nonlinear dimensionality expansion to multispectral imagery
(Chapter 31), and multispectral magnetic resonance imaging
(Chapter 32). Application is not a real strength of this book for
that readers can refer elsewhere. The final summary (Chapter
33) could have been much shorter and more impactful. I
recommend the short conclusions written after each chapter
and also that readers review them more than once after
reading a chapter.
Overall, this is a great book on: (a) hyperspectral image
processing, and (b) hyperspectral signal processing. I recommend
this book as a must read to anyone interested in hyperspectral
remote sensing in general, and hyperspectral image or signal
processing in particular. The book is an excellent reference
material for students, teachers, professionals, and practitioners
throughout remote sensing and related community. I make a
strong recommendation to anyone interested in hyperspectral
image processing, and hyperspectral signal processing to make
this book a common reference.
References
Chang, C.I., 2003. “
Hyperspectral Imaging: Techniques for
Spectral Detection and Classification
”. Kluwer Academic/
Plenum Publishers, Dordrecht, the Netherlands.
Plaza, A., and Chang, C.I. (eds.) 2007. “
High performance
computing in remote sensing
”. CRC Press, Boca Raton, FL.
Thenkabail, P.S., Lyon, G.J., and Huete, A. (eds.) 2012.
“Hyperspectral Remote Sensing of Vegetation”. CRC
Press- Taylor and Francis group, Boca Raton, London,
New York. 781 pp.
