[收藏]直方图熵阈值算法（转）

最新推荐文章于 2024-04-16 11:55:49 发布

cwqcgx

最新推荐文章于 2024-04-16 11:55:49 发布

阅读量3.4k

点赞数 1

CC 4.0 BY-SA版权

分类专栏：图像处理文章标签：算法 object filter image c

本文链接：https://blog.youkuaiyun.com/cwqcgx/article/details/6233525

本文介绍了直方图熵阈值算法在图像分割中的应用，包括一维和二维直方图熵阈值法。通过计算图像像素的熵，寻找最佳阈值进行分割。算法首先计算一维直方图，然后通过最大熵原则确定最优阈值。接着，文章扩展到二维直方图，考虑像素的局部灰度信息，进一步提高分割效果。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

转自http://hi.baidu.com/yibobin/blog/item/3a84c8f7047dee2f720eec81.html

一维直方图熵阈值算法

% 1D entropy thresholding method
% Pun提出，Kapur对其阈值和熵进行改进

% 两类：object 和background
% P1 = sum(pi) i：1~T
% P2 = sum(pi) i:T+1~255
% HO = ln(P1) + H1/P1;
% HB = ln(P2) + H2/P2;
% H1 = -sum(pi*ln(pi)); i:1~T
% H2 = -sum(pi*ln(pi)); i:T+1~255
% H = HO + HB;

% T_best = argmax(H);

clc;
clear;

%I = imread('E:/test/chinalake.bmp','bmp');
I = imread('E:/test/lena.png','png');

I = double(I);

I = Medianfilter(I); % median filter

[height,width] = size(I);
Size = height * width; % the size of the image
h_T = sum(sum(I)); % the total gray value of the image

G_min = min(min(I)); % the min gray value of the image
G_max = max(max(I)); % the max gray value of the iamge

I_seg = zeros(height,width); % the array to store the segmented image
I_hist = zeros(1,256); % the array to store the hist of the image

thresh = 0; % the threshold

num1 = 0;
num2 = 0;            % count the num of the pixel from the diffrient class
P1 = 0;
P2 = 0;              % the probability of the different class
h_T1 = 0;
h_T2 = 0;            % the total gray value of different class

max = 0;

H1 = 0;
H2 = 0;             % the middle var
H_object = 0;
H_background = 0;
H_total = 0;        % the total entropy
T_best = 0;         % the best thresh

%%%%% 计算直方图 %%%%%%
for i=1:height        % calculate the hist of the image
    for j=1:width