Connected Components on Input Image

Hi all

I want to visualize my compnents on original input image. I have performed foregorund detection by some background subtraction algorithm. Let say I have foregorund mask fgmask and input frame I.I have performed the connected component analysis like that

I=input Image;

FgMask=binary Image;

labelImage=bwlabel(FgMask,8);

blobMeasurements = regionprops(labeledImage, I, 'all');

numberOfBlobs = size(blobMeasurements, 1);

Now what I want to do, actually I do not need bounding box on foreground region, I want to label foregorund region into original input Image (foregorund region should be labeled from red colours) instead of bounding box.HOw can I perform that??

Every help is appreciated.

Answer 1

[Image Analyst]

You can tint your image red in the foreground regions

redChannel = grayImage; % Initialize.
redChannel(binaryImage) = 255; % Max out red in foreground pixels.
% Create an RGB image.
tintedImage = cat(3, redChannel, grayImage, grayImage);
imshow(tintedImage);

[Algorithms Analyst]

Thank you very much Dear Image Analyst.

How can I make a bounding box in the resultant red region in input Image after that I mean

tintedImage=cat(3,redChannel,grayImage,grayImage);

Thanks

[Image Analyst]

You can use the plot() function to plot lines between the 4 corners. Or use the rectangle() function.

[Algorithms Analyst]

How can I get the coodinates of the bounding box rectangle from tintedImage?

[Image Analyst]

Originally you said you didn't need it. Anyway, you can just use regionprops

labeledImage = bwlabel(binaryImage);
measurements = regionprops(labeledImage, 'BoundingBox');
allBoundingBoxes = [measurements.BoundingBox];