RollingGuidanceFilter_Guided_导向滤波_滚动导向滤波_

% % Rolling Guidance Filter % % res = RollingGuidanceFilter(I,sigma_s,sigma_r,iteration) filters image % "I" by removing its small structures. The borderline between "small" % and "large" is determined by the parameter sigma_s. The sigma_r is % fixed to 0.1. The filter is an iteration process. "iteration" is used % to control the number of iterations. % % Paras: % @I : input image, DOUBLE image, any # of channels % @sigma_s : spatial sigma (default 3.0). Controlling the spatial % weight of bilateral filter and also the filtering scale of % rolling guidance filter. % @sigma_r : range sigma (default 0.1). Controlling the range weight of % bilateral filter. % @iteration : the iteration number of rolling guidance (default 4). % % % Example % ========== % I = im2double(imread('image.png')); % res = RollingGuidanceFilter(I,3,0.05,4); % figure, imshow(res); % % % Note % ========== % This implementation filters multi-channel/color image by separating its % channels, so the result of this implementation will be different with % that in the corresponding paper. To generate the results in the paper, % please refer to our executable file or C++ implementation on our % website. % % ========== % The Code is created based on the method described in the following paper: % [1] "Rolling Guidance Filter", Qi Zhang, Li Xu, Jiaya Jia, European % Conference on Computer Vision (ECCV), 2014 % % The code and the algorithm are for non-comercial use only. % % % Author: Qi Zhang (zhangqi@cse.cuhk.edu.hk) % Date : 08/14/2014 % Version : 1.0 % Copyright 2014, The Chinese University of Hong Kong. % function res = RollingGuidanceFilter_Guided(I,sigma_s,sigma_r,iteration) if ~exist('iteration','var') iteration = 4; end if ~exist('sigma_s','var') sigma_s = 3; end if ~exist('sigma_r','var') sigma_r = 0.1; end res = gaussFilter(I,sigma_s); for i=1:iteration for c=1:size(I,3) G = res(:,:,c); res(:,:,c) = guidedfilter(G,I(:,:,c),sigma_s,sigma_r^2); end end end