Detecting blob coordinates in a small image

Hi guys!

I need to get the blob coordinates on a very small image (which seems to be an issue) shown below. It’s a portion of a processed camera feed.

as of now i am stuck as i cannot detect the mentioned blob, no matter the parameters
the blob detector call and keypoint drawing

keypointsR = det.detect(rROImask)
highlightR = cv2.drawKeypoints(rROImask, keypointsR,np.array([]),(255,0,0), cv2.DRAW_MATCHES_FLAGS_DRAW_RICH_KEYPOINTS))

Do you have any idea what am i doing wrong?

Detectors analyze the surroundings of each pixel to get a keypoint. They use a sliding window, and won’t work if the window doesn’t fit in the image.

1 Like

What you are doing wrong: not telling how small “very small” is, and not showing an example image.

(Note that if the size is the issue, you can always resize the image…)

can you explain, why you can’t use the whole image ?

I believe the image @Pendzel provided is the example:

@Pendzel , can you tell us the image size, and show more detailed code, like what detector are you using?

i cant use it due to performance issues, Let’s say that’s not the only process running. Alejandro is right it wont work i solved it with detecting the edges instead of a blob.
Case is closed, Tahnks Guys.

1 Like