In Matlab, getting error "TrueColor CData contains element out of range 0.0 <= value <= 1.0"

Question

I am making a function that takes an image and a luminance range and then returns an intensity sliced image. So essentially it detects if a certain pixel's intensity is in the range and then sets the intensity to 255 so it's bright white, or otherwise leaves it alone. The equation for luminance I'm given is L = 0.3R + 0.59G + 0.11B. I think everything is executed correctly, but when I try to use the built in image function to turn the intensities into an image I get the error in the title. Here is my code so far.

function finalImage = nicholson11(image,range)

originalImage = imread(image);
[W,H,clrs] = size(originalImage);
finalImage = zeros(W,H,3);

for i=1:1:W
for j=1:1:H
if ((((originalImage(i,j,1))*0.3) +((originalImage(i,j,2))*0.59) +((originalImage(i,j,3))*0.11)) >=range(1) && ((((originalImage(i,j,1))*0.3) +((originalImage(i,j,2))*0.59) + ((originalImage(i,j,3))*0.11))<=range(2)))
finalImage(i,j,:)=255;

end
end
end

Answer 1

You should consider vectorizing your code to make it more legible and less obfuscated. You can completely avoid for loops by doing the following:

function finalImage = nicholson11(image,range)

imDouble = double(image);

intensityImage = 0.3*imDouble(:,:,1) + 0.59*imDouble(:,:,2) + 0.11*imDouble(:,:,3);
finalImage = zeros(size(intensityImage));
finalImage(intensityImage >= range(1) | intensityImage <= range(2)) = 255;

% Assuming your original image was uint8
finalImage = im2uint8(finalImage);

Try that and see if you get the same results. Also, going with the comments that have already been said, when you create finalImage, this is a double array, which assumes that the intensities are between [0,1]. As such, you will either need to normalize your intensities so that they fall in this range, or use a conversion routine that will get it to the type that you want. Assuming that your image was originally uint8, you can use the im2uint8 command.