1 简介
提出了一种基于模糊熵和FCM的彩色图像聚类分割算法.该算法可以自动确定图像的颜色类数目和初始类中心,从而提高了聚类的收敛速度,并且能够解决模糊熵阈值化分割算法所造成的过度分割问题.首先,计算彩色图像各颜色分量的模糊熵,获得分量模糊熵曲线,并根据模糊熵原理确定各分量的分割区域及聚类中心;然后,对各分量的聚类中心进行组合,形成彩色图像可能的聚类中心.但是,组合的聚类中心数目会多于实际的聚类数目,造成过度分割.
2 部分代码
function [center, U, obj_fcn] = iffcm(data, cluster_n)
if nargin ~= 2 & nargin ~= 3,
error('Too many or too few input arguments!');
end
data_n = size(data, 1);
in_n = size(data, 2);
% Change the following to set default options
default_options = [2; % exponent for the partition matrix U
100; % max. number of iteration
1e-5; % min. amount of improvement
1]; % info display during iteration
if nargin == 2,
options = default_options;
else
% If "options" is not fully specified, pad it with default values.
if length(options) < 4,
tmp = default_options;
tmp(1:length(options)) = options;
options = tmp;
end
% If some entries of "options" are nan's, replace them with defaults.
nan_index = find(isnan(options)==1);
options(nan_index) = default_options(nan_index);
if options(1) <= 1,
error('The exponent should be greater than 1!');
end
end
expo = options(1); % Exponent for U
max_iter = options(2); % Max. iteration
min_impro = options(3); % Min. improvement
display = options(4); % Display info or not
% alfa=0.9;
obj_fcn = zeros(max_iter, 1); % Array for objective function
%U = initfcm(cluster_n, data_n); % Initial fuzzy partition
center= initifcmv(cluster_n); %初始化聚类中心
[histdata,histrate]=datahistprocess(data);%data:图像灰度值除255归一化，hist:灰度级对应的频数
% Main loop
for i = 1:max_iter,
%[U, center, obj_fcn(i)] = fcm_spatial_stepfcm(data,data_spatial, U, cluster_n, expo, beta);
[U, center, obj_fcn(i)] = iffcm_step(histdata, center, cluster_n, expo,histrate);
if display,
fprintf('Iteration count = %d, obj. fcn = %f\n', i, obj_fcn(i));
end
% check termination condition
if i > 1,
% if abs(obj_fcn(i) - obj_fcn(i-1)) < min_impro, break; end,
% if norm(old_U-U,2) < min_impro, break; end,
% if norm(old_U-U,'fro') < min_impro, break; end,
if norm(old_center-center,2) < min_impro, break; end
end
old_U = U;
old_center = center;
end
U=u_return(old_U,data);
iter_n = i; % Actual number of iterations
% obj_fcn(iter_n+1:max_iter) = [];
mf = U.^expo; % 隶属度矩阵进行指数运算结果
dist = distfcm(old_center, data); % 计算距离矩阵
obj_fcn = sum(sum((dist.^2).*mf)); % 计算目标函数值
3 仿真结果
4 参考文献
[1]李桂芝等. "基于模糊熵和RPCL的彩色图像聚类分割." 中国图象图形学报 10.10(2005):5.
博主简介：擅长智能优化算法、神经网络预测、信号处理、元胞自动机、图像处理、路径规划、无人机等多种领域的Matlab仿真，相关matlab代码问题可私信交流。
部分理论引用网络文献，若有侵权联系博主删除。
