readme

脊波工具和相关实验以及有关实验数据

          Finite Ridgelet Transform Toolbox (FRIT) - Version 1.0
          ------------------------------------------------------

FRIT is a Matlab toolbox that implements the finite ridgelet transforms.
The transform algorithms implemented in this toolbox are based on the
following papers:

1. M. N. Do and M. Vetterli, "Orthogonal finite ridgelet transform for image 
compression", IEEE International Conference on Image Processing (ICIP), 
Vancouver, Canada, September 2000.

2. M. N. Do and M. Vetterli, "Image denoising using orthonormal finite 
ridgelet transform", Proc. of SPIE Conf. on Wavelet Applications in Signal 
and Image Processing VIII, San Diego, USA, August 2000.

3. M. N. Do and M. Vetterli, "The finite image transform for image
representation", IEEE Transactions on Image Processing, January 2003.


The following invertible transforms are provided in the toolbox:

FRAT	Finite Radon transform.
FRATC   The MEX file (may need to be compiled before the first use) 
        that does the actual computation of the FRAT.
FRATF  	Finite Radon transform.  This version use FFT2 and is 100% in Matlab 
	but slower than FRAT.
FRIT  	Finite ridgelet transform.  This is the general version of the
	finite ridgelet transform.
FRITO	Orthonormal finite ridgelet transform.  There are additional
	restrictions on the image size and number of wavelet decomposition
	levels.
FRITM	Orthonormal finite ridgelet transform mixed with DCT.  Same
	restriction as for FRITO.
FRATFD	Folded finite Radon transform.
FRITFD	Folded finite ridgelet transform
AFRIT  	Adaptive finite ridgelet transform.
BFRIT   Block finite ridgelet transform (test version!)

The inverse transforms for those are I___, for example IFRIT.  The
script file FRITDEMO contains examples on how to use some of these
functions.  In addition, the files COMPARE_NLA and COMPARE_DENOISE
compare the performance of the FRIT against other traditional transforms
and produce images that included in the above papers. 

This toolbox requires the Wavelet Toolbox by MathWorks.  It has been tested
on the Matlab Version 6.1 and the Wavelet Toolbox Version 2.1.

To use the toolbox, before the first run, the two C programs FRATC and
IFRATC may have to be complied into mex files.  That means, from the
Matlab command window, type:
	mex fratc.c 
	mex ifratc.c

We provide pre-compiled mex files for several platforms.

Comments, suggestions, or questions can be sent to:
[e-mail]
http://www.ifp.uiuc.edu/~minhdo

Swiss Federal Institute of Technology, Lausanne, November 2001.
University of Illinois at Urbana-Champaign, November 2003.