compute_dead_leaves_image.m

image denoising toolbox in matlab

function M = compute_dead_leaves_image(n,sigma,options)

% compute_dead_leaves_image - compute a random image using the dead-leaves model
%
%   M = compute_dead_leaves_image(n,sigma,options);
%
%   n is the size of the image.
%   sigma>0 control the repartition of the size of the basic shape:
%       sigma --> 0  gives more uniform repartition of shape
%       sigma=3 gives a nearly scale invariant image.
%   options.nbr_iter put a bound on the number of iteration.
%   options.shape can be 'disk' or 'square'
%
% References : 
%   Dead leaves correct simulation :
%    http://www.warwick.ac.uk/statsdept/staff/WSK/dead.html
%
%   Mathematical analysis
%    The dead leaves model : general results and limits at small scales
%    Yann Gousseau, Fran竎ois Roueff, Preprint 2003
%
%   Scale invariance in the case sigma=3
%    Occlusion Models for Natural Images: A Statistical Study of a Scale-Invariant Dead Leaves Model
%     Lee, Mumford and Huang, IJCV 2003.
%
%   Copyright (c) 2005 Gabriel Peyr