Introduction - Feature Extraction and Image Processing - page 15

information flow during processing. However, the underlying mathematics is not made
clear to the user, as it can be when a mathematical system is used. There is a new textbook,
and a very readable one at that, by Nick Efford (Efford, 2000) which is based entirely on
Java and includes, via a CD, the classes necessary for image processing software development.
A set of WWW links are shown in Table 1.2 for established freeware and commercial
software image processing systems. What is perhaps the best selection can be found at the
general site, from the computer vision homepage software site (repeated later in Table 1.5 ).

Table 1.2 Software package websites

Packages (freeware or student version indicated by *)

General Site Carnegie Mellon

http://www.cs.cmu.edu/afs/cs/
project/cil/ftp/html/v-source.html

Khoros Khoral Research

http://www.khoral.com/

Hannover U

http://www.tnt.uni-hannover.de/
soft/imgproc/khoros/

AdOculos* The Imaging Source

http://www.theimagingsource.com/

(+ Textbook)

catalog/soft/dbs/ao.htm

CVIPtools* Southern Illinois U

http://www.ee.siue.edu/CVIPtools/

LaboImage* Geneva U

http://cuiwww.unige.ch/~vision/
LaboImage/labo.html

TN-Image* Thomas J. Nelson

http://las1.ninds.nih.gov/tnimage-
manual/tnimage-manual.html

1.5 Mathematical systems

In recent years, a number of mathematical systems have been developed. These offer what
is virtually a word-processing system for mathematicians and many are screen-based using
a Windows system. The advantage of these systems is that you can transpose mathematics
pretty well directly from textbooks, and see how it works. Code functionality is not obscured
by the use of data structures, though this can make the code appear cumbersome. A major
advantage is that the system provides the low-level functionality and data visualisation
schemes, allowing the user to concentrate on techniques alone. Accordingly, these systems
afford an excellent route to understand, and appreciate, mathematical systems prior to
development of application code, and to check the final code functions correctly.

1.5.1 Mathematical tools

Mathcad, Mathematica, Maple and Matlab are amongst the most popular of current tools.
There have been surveys that compare their efficacy, but it is difficult to ensure precise
comparison due to the impressive speed of development of techniques. Most systems have
their protagonists and detractors, as in any commercial system. There are many topics
which use these packages for particular subjects, and there are often handbooks as addenda
to the packages. We shall use both Matlab and Mathcad throughout this text as they are