Avoid for loop to increase processing in image processing.

I have three matrices

labelism is nr x nc binary matrix

colorim is nr x nc x 3 RGB matrix

and skin_img is also nr x nc x 3 RGB matrix with initially all the zeros on it (black image)

I want to use the labelism matrix to create skin_img matrix from colorim matrix.

Following is the code giving perfectly fine answer

 for i = 1:nr
     for j = 1:nc
         if (labelim(i,j) == 1)
             skin_img(i,j,:) = colorim(i,j,:);
         end
     end
 end
 
 imshow(uint8(skin_img));

But it is slow for big size images. Is there any way to avoid for loop?

Answer 1

[Alex Mcaulley]

It is just masking your RGB image. Isn't it?

skin_img = repmat(labelism,[1,1,3]).*colorim;

[Muhammad Farhan Mughal]

Thank you Alex, yes it was about masking. The processing speed has increased a lot.

Thanks again