How to use the code:

This package is self-contained, to implement multi-Wiener SURE-LET deconvolution algorithm. 
The main algorithm file is: SURE_LET_Deconvolution.m


After changing MATLAB into the current directory, you may start to use it in two ways:
[1] run demo.m file, you can change the experiment setting (images, blurs and noise levels);
[2] run SURE_LET_GUI for GUI demo.


Key experimental settings:
[1] you can copy your test image into this folder, change the image_name in demo.m;
[2] you can add new blur types in Image_blurring.m;
[3] you can change noise level in demo.m, where nsigma refers to noise standard deviation. 


Reference
[1] F. Xue, F. Luisier and T. Blu,"SURE-LET image deconvolution using multiple Wiener filters", Proceedings of the 2012 IEEE International Conference on Image Processing (ICIP'12), Orlando, USA, pp. 3037-3040, September 30--October 3, 2012. 
[2] F. Xue, F. Luisier and T. Blu, "Multi-Wiener SURE-LET Deconvolution", IEEE Transactions on Image Processing, Vol. 22 (5), pp. 1954-1968, May 2013. 

Contact: Thierry Blu, the Chinese University of Hong Kong, Shatin, Hong Kong
mailto:thierry.blu@m4x.org