How can I apply a non-built in function to an image for quantification?

THIS ERROR APPEARS : Assignment has more non-singleton rhs dimensions than non-singleton subscripts Error in TP2_script (line 13) I(i,j)=quant;

Answer 1

[Walter Roberson]

 quant= [I(i,j)*(fmax-fmin)/q]/N-1;

[oumi k]

Thank you for your answer. Apparently I now have a new error with this line in the function

Error using Quant_gray --> if I(i,j)~= fmax; %I(i,j) not equal to fmax

Not enough input arguments.

Can you please advise?

[oumi k]

The image is all entirely black even with this modification.

Answer 2

[Image Analyst]

Try this. You might want to reexamine your formula though.

% I=imread('irm2.jpg');
grayImage=imread('cameraman.tif');
[rows, columns, numberOfColorChannels] = size(grayImage);
subplot(1, 2, 1);
imshow(grayImage);
title('Input Image')
impixelinfo
axis('on', 'image')
if numberOfColorChannels == 3
    errordlg('Error: program only works for gray scale images, not color images');
    return
end
N=16 %niveau de quantification change to 4,8,16...128
fmin=min(min(grayImage)) % for 2D matrix of data (vector utilisation)
fmax=max(max(grayImage))
q=(fmax-fmin)/N % pas de quantification
% Convert to floating point
grayImage = double(grayImage);
for row = 1 : rows
    for col = 1 : columns
        quant = Quant_gray(grayImage, N, q, fmax, fmin, row, col);
        grayImage(row, col) = quant;
    end
end
% Show resulting output image.
subplot(1, 2, 2);
imshow(grayImage, []);
title('Output Image')
impixelinfo

%===================================================================================================================
function [quant] = Quant_gray(I, N, q, fmax, fmin, row, col)
if I(row, col) ~= fmax %I(i,j) not equal to fmax
    quant= [I(row, col) * (fmax-fmin)/q] / N - 1;
    quant= floor(quant);
else
    quant = N-1; % I(i,j)=fmax
end
end

[oumi k]

Thank you for your answer. The image output is entirely black though with the code you provided me with. The goal is for me to use the formula as an algortihm for the image quantification, however I am not sure where I went wrong.

[Image Analyst]

I suspect this

quant= [I(row, col) * (fmax-fmin)/q] / N - 1;
quant= floor(quant);

is always giving 0. Not sure what you're doing but you might try discretize().

[oumi k]

Yes the problem is here. Thank you for your advice I will work on it.