How to Update Image with Maximum Connected Component of pixel?

Step 1: Extract Conncted Component, CC.

Step 2: Idx=max(CC)

Step3: I=I(Idx) / Update Image with Maximum Connected Component of pixel

Please anyone help me, I can't understant this.

Answer 1

[Image Analyst]

This does not make any practical sense. You wouldn't want to label the image (give an ID number to every blob) and then take the ID number that is greatest. Yes, you could do it, and it would take (roughly) the right-most, bottom-most blob, but that is not the correct way to get that blob. Blob are labeled in a top/down, left-to-right fashion as they are encountered by scanning the image in function bwlabel() or bwconcomp().

On the other hand, if you're trying to get the blob with the max area, instead of the max label (ID number), then you'd want bwareafilt():

largestBlobOnly = bwareafilt(binaryImage, 1); % Extract largest blob.

Explain in more detail what you want to do so I can know how to answer you.

[Nadim Kaysar]

This is the algorithm that I want to implement. I attached the algorithm. You can see it.

[Image Analyst]

Not sure what steps 2 and 3 mean but I suspect it's to extract the largest blob, but maybe it means the dilation (local max, done by imdilate() function). Read the rest of the paper or ask the author to determine exactly what those steps mean.