How to plot RGB histogram of an image into a single 3D slice plot?

Hello. Im working on color image encryption and need to plot the histogram for each color channels. I can plot the individual color channel histogram in three 2D plot with this code.

close all;
image = imread('peppers.png');
R = image(:,:,1);
G = image(:,:,2);
B = image(:,:,3);
[Width,Length] = size(image);
    subplot(2,2,1); imshow(image);
        title('Plain Image')
    subplot(2,2,2); 
        imhist(R);
        myHist1 = findobj(gca, 'Type', 'Stem');
        myHist1.Color = [1 0 0];
        title('Red Channel Plain Histogram')
    subplot(2,2,3); 
        imhist(G);
        myHist2 = findobj(gca, 'Type', 'Stem');
        myHist2.Color = [0 1 0];
        title('Green Channel Plain Histogram')
    subplot(2,2,4);
        imhist(B);
        myHist3 = findobj(gca, 'Type', 'Stem');
        myHist3.Color = [0 0 1];
        xlim([0 256]);
        title('Blue Channel Plain Histogram')

But, i tried to plot them together in a 3D plot like this below (picture from an article i found), but i cant. I tried to find any forum that talks about this, but i can only found slice plots of meshgrid, not from histogram.

If anyone could help, I would appreciate it so much.

Thank you!

Answer 1

[Voss]

image = imread('peppers.png');
R = image(:,:,1);
G = image(:,:,2);
B = image(:,:,3);

subplot(2,2,1);
imshow(image);
title('Plain Image')

subplot(2,2,2); 
imhist(R);
myHist1 = findobj(gca, 'Type', 'Stem');
myHist1.Color = [1 0 0];
xlim([0 255]);
title('Red Channel Plain Histogram')

subplot(2,2,3); 
imhist(G);
myHist2 = findobj(gca, 'Type', 'Stem');
myHist2.Color = [0 1 0];
xlim([0 255]);
title('Green Channel Plain Histogram')

subplot(2,2,4);
imhist(B);
myHist3 = findobj(gca, 'Type', 'Stem');
myHist3.Color = [0 0 1];
xlim([0 255]);
title('Blue Channel Plain Histogram')

figure('Position',[10 10 500 500])
ax = gca();

myHist1_new = copyobj(myHist1,ax);
myHist1_new.ZData = myHist1_new.YData;
myHist1_new.YData = 1+zeros(size(myHist1_new.XData));

myHist2_new = copyobj(myHist2,ax);
myHist2_new.ZData = myHist2_new.YData;
myHist2_new.YData = 2+zeros(size(myHist2_new.XData));

myHist3_new = copyobj(myHist3,ax);
myHist3_new.ZData = myHist3_new.YData;
myHist3_new.YData = 3+zeros(size(myHist3_new.XData));

box on
grid on
xlim([0 255])
% ax.XDir = 'reverse'; % to match the xdir in the reference image

ylim([0 4])
yticks([1 2 3])
yticklabels(["R","G","B"]+" Channel")

view([30 35])

[Mohammad]

It works like a charm! I only have MATLAB 2014b though which yticks is yet undefined, but i successfully modify and run it.

Thank you!

[Voss]

You're welcome!

Answer 2

[Ashutosh Thakur]

Hello Mohammad,

You can explore the hist3 function available in MATLAB to create 3-D histogram plots. However, note that this function is specifically designed for bivariate data. You can refer to following link for the usage and the examples of the hist3 function: https://www.mathworks.com/help/stats/hist3.html.

You can also refer to the sample code for the usage of hist3 function:

x = randn(1000, 1); % 1000 random values from a normal distribution
y = randn(1000, 1); % 1000 random values from a normal distribution

% Combine the data into a single matrix
data = [x, y];

% Create a 3-D histogram plot
figure;
hist3(data, 'CdataMode', 'auto', 'FaceColor', 'interp');

% Set plot properties
title('3-D Histogram Plot');
xlabel('X-axis values');
ylabel('Y-axis values');
zlabel('Frequency');
colorbar; % Add a color bar to show the frequency scale

% Set view angle for better visualization
view(3);
grid on;

I hope this helps!