Retrieving the lung pixels through matrix operations with images

I have two images, an original CXR, and its corresponding lung mask. I would like to perform an operation with these two images, to generate only the lung ROI from the original CXR and make the rest of the background black. The resultant image should be in RGB as the original CXR and mask, and not grayscale or binary. I tried performing bitwise multiplications and X-OR operations but doesn't work. Can you suggest me the code to do this?

>>

[KALYAN ACHARJYA]

I have edited the answer, Have you check? Please confirm?

Answer 1

[KALYAN ACHARJYA]

%This can do without for loop also, recomemded to follow the without loop  
[rows colm]=size(rgb_image);
      for i=1:rows
          for j=1:colm
              if (mask(i,j)==0) %Mask should be binary Image, if not convert it.
                  rgb_image(i,j,1)=0;
                  rgb_image(i,j,2)=0;
                  rgb_image(i,j,3)=0;
              end
          end 
      end 
      figure,imshow(rgb_image);

[Sivaramakrishnan Rajaraman]

it doesn't work. can you be a bit more clearer on this? thanks.

[KALYAN ACHARJYA]

• Another way using if else-
• if mask image pixel (position) is black
• force RGB image to black
• else
• No change in RGB Image

[KALYAN ACHARJYA]

I Have no time now, by late eve (10 PM) I will reply to you with detail code.

[Sivaramakrishnan Rajaraman]

Many thanks and looking forward

[Sivaramakrishnan Rajaraman]

Do you mind helping with your code in this regard? many thanks.

[Image Analyst]

Maybe the single line of code will be easier for you to understand since it's basically just multiplication of the image by the mask:

maskedRgbImage = bsxfun(@times, maskedRgbImage, cast(mask, 'like', maskedRgbImage));

[KALYAN ACHARJYA]

Yes sir @Image Analyst efficient too, I have tried from my side, unable to execute in that time, that's why I have given the loop options with a disclaimer

[Image Analyst]

Hmmmm. Not sure why. Here's a fully working example.

originalImage = imread('peppers.png');
subplot(2, 2, 1);
imshow(originalImage);
axis('on', 'image');
[rows, columns, numberOfColorChannels] = size(originalImage);
% Make mask
mask = false(rows, columns);
mask(100:300, 100:400) = true;
subplot(2, 2, 2);
imshow(mask);
axis('on', 'image');

maskedRgbImage = originalImage; % Initialize
% Mask the image using bsxfun() function to multiply the mask by each channel individually.
% Will work regardless if it's grayscale or color.
maskedRgbImage = bsxfun(@times, maskedRgbImage, cast(mask, 'like', maskedRgbImage));
subplot(2, 2, 3);
imshow(maskedRgbImage);
axis('on', 'image');

[Sivaramakrishnan Rajaraman]

It works pretty well for me. Many thanks.

Answer 2

[Image Analyst]

This should work

maskedRgbImage = originalImage; % Initialize
% Convert to RGB, if needed.
[rows, columns, numberOfColorChannels] = size(maskedRgbImage);
if numberOfColorChannels == 1
    maskedImage = cat(3, maskedRgbImage, maskedRgbImage, maskedRgbImage);
end
% Mask the image using bsxfun() function to multiply the mask by each channel individually.
% Will work regardless if it's grayscale or color.
maskedRgbImage = bsxfun(@times, maskedRgbImage, cast(mask, 'like', maskedRgbImage));