The Spectral Signal Processing Suite (SSPS), version 1.0, is for the
of increasing the accuracy of discrete Chebyshev approximations of
discontinuous functions by reducing or eliminating the Gibbs-Wilbraham
phenomenon. The suite may be used to implement a postprocessing method
Chebyshev Collocation (Pseudospectral) or Chebyshev Spectral Viscosity
for solving Partial Differential Equations. The suite implements
filtering, edge detection, and Gegenbauer Reconstruction postprocessing
The Spectral Signal Processing Library is written the Java
language and has a graphical user interface which conveniently wraps
numerical classes and provides a graphical examination of the edge
and postprocessing process. All source code is freely available and the
may be used in any way that is desired as long as links/references to
are clearly provided in any work which uses the library or in any
incorporates or extends the Spectral Signal Processing Suite. The
Signal Processing Suite comes with absolutely no warranty of any kind.
reports and coding suggestions are welcome and should be sent to
the algorithms implemented in the software and examples using the
be found in the paper The
Signal Processing Suite which appears in the
Mathematical Software, vol. 29, no. 2, June 2003.
1024x768 screen size). Here is a newer applet running
which now implements edge detection and postprocessing for Fourier
Download signal1_0.jar and place in
of your choice, e.g. c:\jp\signal1_0.jar. Modify your classpath
find the package. At the DOS or shell prompt enter set
Set up a directory for your application, say c:\j\postProcess.
pp.java so that the setDataDirectory( ) method sets the data
directory to the directory of your choice. To run the application type java
-cp signal1_0.jar pp at the prompt.