lowpass.m

function [sl, sh] = lowpass(s, lambda, npad)

% lowpass -- Lowpass filter image and return low and high frequency
%            components, consisting of the lowpass filtered image and
%            its difference with the input image. The lowpass filter
%            is equivalent to Tikhonov regularization with lambda as
%            the regularization parameter and a discrete gradient as
%            the operator in the regularization term.
%
% Usage:
%       [sl, sh] = lowpass(s, lambda, npad)
%
% Input:
%       s         Input image or 3d array of images
%       lambda    Regularization parameter controlling lowpass filtering
%       npad      Number of samples to pad at image boundaries
%  
% Output:
%       sl        Lowpass component
%       sh        Highpass component
%
%   
% Author: Brendt Wohlberg <brendt@lanl.gov>  Modified: 2015-04-09
%
% This file is part of the SPORCO library. Details of the copyright
% and user license can be found in the 'License' file distributed with
% the library.


if nargin < 3,
  npad = 16;
end

grv = [-1 1];
gcv = [-1 1]';
Gr = fft2(grv, size(s,1)+2*npad, size(s,2)+2*npad);
Gc = fft2(gcv, size(s,1)+2*npad, size(s,2)+2*npad);
A = 1 + lambda*conj(Gr).*Gr + lambda*conj(Gc).*Gc;
sp = padarray(s, [npad npad], 'symmetric', 'both');
slp = ifft2(bsxfun(@rdivide, fft2(sp), A), 'symmetric');
sl = slp((npad+1):(size(slp,1)-npad), (npad+1):(size(slp,2)-npad), :);
sh = s - sl;

return