Finding the mean of a histogram

How can i find the mean of the histogram(R)?

clear;
im = imread('folders01.jpg');
figure(1); imshow(im); axis on;

[xP, yP] = ginput(4);
xP(5) = xP(1);
yP(5) = yP(1);

[x,y] = meshgrid(1:size(im,2),1:size(im,1));

xC = mean(xP); yC = mean(yP);

for i = 1:4
    m = (yP(i+1) - yP(i)) / (xP(i+1) - xP(i));
    b = yP(i) - xP(i) * m;

    for j = 1:3
        mask = im(:,:,j);
        if(yC > xC * m + b)
            mask(y<x*m+b) = 0;
        else
            mask(y>x*m+b) = 0;
        end

        im(:,:,j) = mask;
    end
end

figure(2);
imshow(im);

R = im(:,:,1)
R = R(R ~= 0); R = R.'; 
figure(3);
subplot(3,1,1)
histogram(R)
xlim([0 260])
set(get(gca,'children'),'facecolor',[1 0 0])
set(get(gca,'children'),'edgecolor',[1 0 0])

Answer 1

[Thiago Henrique Gomes Lobato]

What do you mean by mean? The mean number of bins or the mean of the variables that the histogram represents? For the second you can just take the mean of R. For the first look at histogram properties and average the "values" property.

Answer 2

[Image Analyst]

The mean of the histogram will not be as accurate as the mean of the image since it's quantized into bins. Here's an illustration:

grayImage = imread('cameraman.tif');
h = histogram(grayImage, 16) % 16 bins
grid on;
meanImageGL = mean2(grayImage)
grayLevels = h.BinEdges(1:end-1);
counts = h.Values;
meanBinnedGrayLevel = sum(grayLevels .* counts) / sum(counts)
xline(meanImageGL, 'Color', 'g', 'LineWidth', 2);
xline(meanBinnedGrayLevel, 'Color', 'r', 'LineWidth', 2);
% Or you could use the gray levels at the centers of the bins
centerBinGrayLevels = (h.BinEdges(1:end-1) + h.BinEdges(2:end)) / 2;
meanBinnedGrayLevel2 = sum(centerBinGrayLevels .* counts) / sum(counts)

You'll see

meanImageGL =

118.724487304688

meanBinnedGrayLevel =

110.93896484375

meanBinnedGrayLevel2 =

118.93896484375

So, which do you want?