How do i find the RGB values of color image in range 1 to 255

I want know the RGB values are in the form of three digit like 100,243,122,etc.

please anyone can tell me how to find RGB values....

[David Hill]

What is the input?

[DGM]

How is this different from the other question?

https://www.mathworks.com/matlabcentral/answers/1867463-i-have-three-rgb-values-i-want-to-know-color-of-this-rgb-values-how-can-i-do-this

... because I still don't see that the questions have been resolved. So far, you've asked for

• the color of RGB values
• the RGB values of color
• the RGB values of RGB values? images??

What information do you have?

What information do you want?

If you're going to say "I want the RGB value of an image", which one? The value at a specific location? The local average at a specific location? The overall average color? The most frequent color?

[MarKf]

Maybe, since in that previous linked question the answer showing the color from the RGB values was accepted, that is:

rows = 5;
columns = 6;
R = 100 * ones(rows, columns, 'uint8');
G = 243 * ones(rows, columns, 'uint8');
B = 122 * ones(rows, columns, 'uint8');
rgbImage = cat(3, R, G, B);
imgrgb = imshow(rgbImage);

Now the reverse is requested?

You can find the RGB values from any created and read image by selecting the CData of that pixel:

squeeze(imgrgb.CData(1,1,:))'

ans = 1×3

100 243 122

[Shri.s]

Sir i want know RGB values of any color image

[DGM]

As Image Analyst mentioned in the prior thread, and as I mentioned in the comment above and in the answer below,

If you're going to say "I want the RGB value of an image", which one?

• The value at a specific location?
• The local average at a specific location?
• The overall average color?
• The most frequent color?
• All the colors?
• The unique colors?

So far, Image Analyst and Benjamin have given examples of applying basic indexing to extract the color at a specific pixel. I've given examples for all of the other cases. If it's none of those, what is it?

[Shri.s]

The most frequent color?

[DGM]

See this thread:

https://www.mathworks.com/matlabcentral/answers/65055-dominant-color-for-an-rgb-image

I offer some simple examples, and @Image Analyst offers one in the comments as well. If either are of interest, you should read the discussion under ImageAnalyst's answer as well. The "most common color" is the maximizing position in a 3-dimensional histogram, but there are often a number of broad, similarly-prominent maxima in the distribution. The global maxima is often not the one that is perceived to be visually important, as color is often correlated to object content.

If the goal is to use color frequency information, it might need to be paired with other forms of analysis, depending on what you hope color frequency represents within the images under test.

Answer 1

[Benjamin Thompson]

If the image is stored in MATLAB as a three dimensional array, which is the usual, then the red/green/blue channel is the third dimension. Sometimes there is a fourth channel alpha for transparency. You can access the entire red channel of such an image using, for example:

redData = myRGBImage(:,:,1);

Or access the red component of pixel 64, 128 specifically:

redDataOnePixel = myRGBImage(64,128,1);

Hope this helps. For a better answer attach a sample image and ask a more detailed question.

[John D'Errico]

The 4 channels of an image may also be CMYK code values, and I recall devices with 6 or even different 7 color inks, used to increase the color gamut for a corresponding device or other purposes. For example, you might have cyan, magenta, yellow, black, orange, green, light grey.

Answer 2

[DGM]

I want to find the RGB color at/around a specific location in an image

That can be done, but it depends how you want to select the points and whether you want the color of a specific pixel or not. Getting the color of a specific pixel is often pretty useless unless you're dealing with synthetic images or you only care about one pixel.

% inputs
inpict = imread('peppers.png');
sampleradius = 5;

% show the image
imshow(inpict);

% pick points somehow.  
% in this case, do it interactively with the mouse; hit ENTER when done
[x y] = getpts(gca);

% get rid of points that are outside the image area
sz = size(inpict);
points = round([x y]);
badpts = any(points<1,2) | points(:,2)>sz(1) | points(:,1)>sz(2);
points = points(~badpts,:);

% get selected tuples
npoints = size(points,1);
pickedcolors = zeros([npoints size(inpict,3)],class(inpict));
if sampleradius > 0
    xx = 1:sz(2);
    yy = (1:sz(1)).';
end
for p = 1:npoints
    x = points(p,1);
    y = points(p,2);
    if sampleradius <= 0
        % just one pixel
        pickedcolors(p,:) = inpict(y,x,:);
    else
        % find the average of pixels within a given distance
        mask = ((xx-x).^2 + (yy-y).^2) <= sampleradius^2;
        samplesize = sum(mask(:));
        for c = 1:size(inpict,3)
            wpict = inpict(:,:,c);
            pickedcolors(p,c) = sum(wpict(mask))/samplesize;
        end
    end
end

% list the colors as a uint8 color table
pickedcolors
% show the color table as a stripe image
imshow(permute(pickedcolors,[1 3 2]))

There are purpose-built color picker tools that can do this more conveniently, but we'll cross that bridge if we come to it.

Also, bear in mind that most things in MATLAB that accept color tuples or color tables only accept them as unit-scale floating point arguments. You can use

pickedcolors = im2double(pickedcolors);

to get results that can be used for those cases.

I want all the colors of an RGB image as a color table

It doesn't really make much sense to get all the colors of an image. That's just a reshaped image. It might be useful to get all the unique colors.

% inputs
inpict = imread('peppers.png');

% get all unique colors
CT = reshape(inpict,[],size(inpict,3)); % this is all the colors
CT = unique(CT,'rows'); % just the unique colors

I want the average color of an RGB image

See also this thread.

meancolor = uint8(reshape(mean(inpict,[1 2]),1,[])) % the average color

I want the color name of an RGB tuple -OR- RGB tuple for a color name

See Stephen's colornames() as mentioned in the prior thread for both. See also the references that @Image Analyst provided.