beamletrdp.m

beamlet变化的工具箱

%  BeamletRDP -- Use Recursive Dyadic Partitioning to extact most `significant'
%                beams.
%   Usage
%     [btree,vtree,stree] = BeamletRDP(mEC);
%   Input
%     mEC     modified beamlet coefficients - different values but same structure,
%             (4n)*(4n)*(log2(n)+1) array
%   Outputs
%     btree   basis tree -- type of best split  [2n-1 * 2n-1]
%     vtree   value tree -- value of best split [2n-1 * 2n-1]
%     stree   split tree -- direction of split  [2n-1 * 2n-1 * 2]
%   Description
%     maximize_{P} sum_{e in P} M(e) - lambda * K(e),
%     where P is a set of all possible dyadic partitioning, 
%     e is a subsquare, M(e) is the largest beamlet transform in the square e,
%     K(e) is a penalty function, in our case, we choose K(e) = sqrt(|e|). 
%     lambda is a constant. 
% 
%     This algorithm is designed to extract the most significant beamlets in 
%     a noisy image. 
%
%
% Part of BeamLab Version:200
% Built:Friday,23-Aug-2002 00:00:00
% This is Copyrighted Material
% For Copying permissions see COPYING.m
% Comments? e-mail [e-mail]
%
%
% Part of BeamLab Version:200
% Built:Saturday,14-Sep-2002 00:00:00
% This is Copyrighted Material
% For Copying permissions see COPYING.m
% Comments? e-mail [e-mail]
%