matlab里颜色直方图的画法

迷南。 2022-08-13 13:36 369阅读 0赞

1.三个颜色的直方图画在一起。

I=imread('sample.bmp');  % 文件名自己改
    siz=size(I);
    I1=reshape(I,siz(1)*siz(2),siz(3));  % 每个颜色通道变为一列
    I1=double(I1);
    [N,X]=hist(I1, [0:1:255]);    % 如果需要小矩形宽一点，划分区域少点，可以把步长改大，比如0:5:255
    bar(X,N(:,[3 2 1]));    % 柱形图，用N(:,[3 2 1])是因为默认绘图的时候采用的颜色顺序为b,g,r,c,m,y,k，跟图片的rgb顺序正好相反，所以把图片列的顺序倒过来，让图片颜色通道跟绘制时的颜色一致
    xlim([0 255])
    hold on
    plot(X,N(:,[3 2 1]));    % 上边界轮廓
    hold off

2.分开画

i=imread('d:\1.jpg');%读取你要看的图像
    R=i(:,:,1);%把RGB各个分量提出
    G=i(:,:,2);
    B=i(:,:,3);
    figure;%画出直方图
    subplot(1,3,1),imhist(R),title('红色');
    subplot(1,3,2),imhist(G),title('绿色');
    subplot(1,3,3),imhist(B),title('蓝色');
    
    你也可以直接右击i选open selection查看里面的像素值。

3.hsv量化

在图像处理技术领域，通常分析彩色图像是对RGB模式下各分量进行分析。如果要进行颜色识别，利用RGB各分量的组合进行分析图像的颜色就比较困难了，所以需要将彩色图像从RGB模式下转换到HSV模式（维基百科）下，分析图像颜色，并设计出颜色分布的直方图，并重新转换到RGB模式下进行显示。本算法是在matlab环境下实现的。具体代码如下：

%%================================

clear  
clc  
close all  
Image = imread('Test3.jpg');  
\[M,N,O\] = size(Image);  
\[h,s,v\] = rgb2hsv(Image);

H = h; S = s; V = v;  
h = h\*360;  
%将hsv空间非等间隔量化：  
% h量化成16级；  
% s量化成4级；  
% v量化成4级；  
for i = 1:M  
for j = 1:N  
if h(i,j)<=15||h(i,j)>345  
H(i,j) = 0;  
end  
if h(i,j)<=25&&h(i,j)>15  
H(i,j) = 1;  
end  
if h(i,j)<=45&&h(i,j)>25  
H(i,j) = 2;  
end  
if h(i,j)<=55&&h(i,j)>45  
H(i,j) = 3;  
end  
if h(i,j)<=80&&h(i,j)>55  
H(i,j) = 4;  
end  
if h(i,j)<=108&&h(i,j)>80  
H(i,j) = 5;  
end  
if h(i,j)<=140&&h(i,j)>108  
H(i,j) = 6;  
end  
if h(i,j)<=165&&h(i,j)>140  
H(i,j) = 7;  
end  
if h(i,j)<=190&&h(i,j)>165  
H(i,j) = 8;  
end  
if h(i,j)<=220&&h(i,j)>190  
H(i,j) = 9;  
end  
if h(i,j)<=255&&h(i,j)>220  
H(i,j) = 10;  
end  
if h(i,j)<=275&&h(i,j)>255  
H(i,j) = 11;  
end  
if h(i,j)<=290&&h(i,j)>275  
H(i,j) = 12;  
end  
if h(i,j)<=316&&h(i,j)>290  
H(i,j) = 13;  
end  
if h(i,j)<=330&&h(i,j)>316  
H(i,j) = 14;  
end  
if h(i,j)<=345&&h(i,j)>330  
H(i,j) = 15;  
end  
end  
end  
for i = 1:M  
for j = 1:N  
if s(i,j)<=0.15&&s(i,j)>0  
S(i,j) = 1;  
end  
if s(i,j)<=0.4&&s(i,j)>0.15  
S(i,j) = 2;  
end  
if s(i,j)<=0.75&&s(i,j)>0.4  
S(i,j) = 3;  
end  
if s(i,j)<=1&&s(i,j)>0.75  
S(i,j) = 4;  
end  
end  
end  
for i = 1:M  
for j = 1:N  
if v(i,j)<=0.15&&v(i,j)>0  
V(i,j) = 1;  
end  
if v(i,j)<=0.4&&v(i,j)>0.15  
V(i,j) = 2;  
end  
if v(i,j)<=0.75&&v(i,j)>0.4  
V(i,j) = 3;  
end  
if v(i,j)<=1&&v(i,j)>0.75  
V(i,j) = 4;  
end  
end  
end

% 构建4\*16二维数组存放H-S数据  
Hist = zeros(16,4);  
for i = 1:M  
for j = 1:N  
for k = 1:16  
for l = 1:4  
if l==S(i,j)&& k==H(i,j)+1  
Hist(k,l) = Hist(k,l)+1;  
end  
end  
end  
end  
end  
for k = 1:16  
for l =1:4  
His((k-1)\*4+l) = Hist(k,l);%转化为一维数组  
end  
end  
His = His/sum(His)\*1000;  
% 手工绘制彩色图像直方图  
% hist\_h  
m=0;  
for j = 1:300  
if rem(j,16)==1 && m<16  
for k = 0:15  
for i = 1:200  
hist\_h(i,j+k) = m;  
end  
end  
m = m+1;  
end  
end  
% hist\_s  
m=0;  
for j = 1:300  
if rem(j,4) == 1 && m<64  
n = rem(m,4);  
for k = 0:3  
for i =1:200  
hist\_s(i,j+k) = n+1;  
end  
end  
m = m+1;  
end  
end  
% hist\_v  
for j = 1:256  
for i = 1:200  
hist\_v(i,j) = 0.98;  
end  
end  
% 把His赋值给hist\_v  
for k = 1:64  
for j = 1:256  
if floor((j-1)/4) == k  
for i = 1:200  
if i<200-His(k+1)%i>His(k+1)%  
hist\_v(i,j) = 0;  
end  
end  
end  
end  
end

%将h、s、v分量图合并转化为RGB模式

I\_H = hsv2rgb(hist\_h/16,hist\_s/4,hist\_v);

% 画图显示  
figure;  
subplot(3,2,1),imshow(Image),title('原图');  
subplot(3,2,2),imshow(H,\[\]),title('H分量图');  
subplot(3,2,3),imshow(S,\[\]),title('S分量图');  
subplot(3,2,4),imshow(V,\[\]),title('V分量图');  
subplot(3,2,5),imshow(I\_rgb,\[\]),title('色彩量化后的RGB图像');  
subplot(3,2,6),imshow(I\_H,\[\]),title('H-S直方图');  
figure,imshow(I\_H);

%%=======================================

对H、S和V的非均匀量化的划分方法有很多种。而且颜色分布直方图中所有分量V的所有值都设置为固定的一个参数。对于彩色图像的灰色部分没有做针对性处理。