![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/175577/image.jpeg)
Extracting blue objects of interest
12 views (last 30 days)
Show older comments
Hi,
I have an image of a circuit board with very blue chips. I want to display the blue chips as blue and display the rest of the board in greyscale.
My approach so far is:
image = imread('board.tif');
redBand = image(:,:,1); greenBand = image(:,:,2); blueBand = image(:,:,3);
After this I want to get everything in greyscale but display the blue chips on the same image. I think I have to put a red and green mask but I'm unsure how to do this. Any insights?
0 Comments
Answers (2)
Christiaan
on 24 Mar 2015
Dear Izuru,
I made a code for you, where first the color 'blue' is detected at a certain threshold. (here 200) Then with a for loop, all points where the color is below the threshold are neglected. The image I used has been put in the attachment.
clc; clear all; close all;
% Read standard MATLAB demo image.
rgbImage = imread('board.jpg');
% Display the original image.
subplot(1, 3, 1);
imshow(rgbImage);
title('Original RGB Image');
% Maximize figure.
set(gcf, 'Position', get(0, 'ScreenSize'));
% Split into color bands.
redBand = rgbImage(:,:, 1);
greenBand = rgbImage(:,:, 2);
blueBand = rgbImage(:,:, 3);
% Display them.
subplot(1, 3, 2);
imshow(blueBand);
title('Blue Band');
% here objects which are parly blue make totaly blue
for i=1:152
for j=1:229
if blueBand(i,j)>200
redBand(i,j)=0.1;
greenBand(i,j)=0.1;
blueBand(i,j)=200;
end
end
end
% here objects which are not blue to dark
for i=1:152
for j=1:229
if blueBand(i,j)<200
redBand(i,j)=redBand(i,j)*0.2;
greenBand(i,j)=greenBand(i,j)*0.2;
blueBand(i,j)=blueBand(i,j)*0.2;
end
end
end
% Display them.
subplot(1, 3, 3);
im = cat(3,redBand,greenBand,blueBand);
imshow(im);
title('Blue Objects');
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/175577/image.jpeg)
A second possibility would be using a tool from the File Exchange. A nice feature is 'imoverlay'. If you call this tool by writing in the prompt : >>imoverlay_tool and then use as a background 'blueband' (0-255), in the foreground'blueband' (200-250) and in the overlay (alpha 0.8) you obtain a figure that is blue where the objects are blue and the rest is gray.
Good luck! Christiaan
0 Comments
Image Analyst
on 24 Mar 2015
You can use my demos that find specified colors. Try the delta E one, or the one that does it in HSV color space.
Change the thresholds in the HSV one to get blue instead of yellow. For the delta E demo, you don't need to do that because you give it an example of the color you want to find by tracing a small region.
See Also
Products
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!