Simple blob detector won't detect big blobs

Question

I'm having issues with Simple Blob Detector not detecting blobs bigger than 300.

I have this image with a Lego block inside of it, but no matter what area I give it it just won't detect it. Any ideas? I'm trying to get the color of the lego, but I want to filter out everything else

params = cv2.SimpleBlobDetector_Params()
params.filterByArea = True
params.minArea = 500 # Tried using from 300 - 3000 and can't get a detection
params.filterByCircularity = False
params.filterByColor = False
params.filterByConvexity = False
params.filterByInertia = False
detector = cv2.SimpleBlobDetector_create(params)
keypoints = detector.detect(img)

Perhaps you need to read in your input image as grayscale. See also https://stackoverflow.com/questions/8076889/how-to-use-opencv-simpleblobdetector — fmw42, Mar 18 '19 at 05:25
@fmw42 BGR images are automatically converted to grayscale, so no need to explicitly do this unless you want to detect in a particular channel. — Piglet, Mar 18 '19 at 09:52
`@Piglet` That does not make sense to me. What command does that? When reading a color image, the reading command must specify to convert to grayscale otherwise all color images would become grayscale. — fmw42, Mar 18 '19 at 17:52

score 0 · Answer 1 · answered Mar 18 '19 at 09:46

Your parameters are insufficient. It's never a good idea to search anything. You should always add as many constraints as possible.

I suggest you add more paramters. I found your brick by simply adding a maxArea of 100000.

A minArea of 500 doesn't make too much sense to me as the smallest lego parts would exceed this by far.

Simple blob detector won't detect big blobs

1 Answers1