- Binarize the image so that it's logical. binaryImage = grayImage > 128
- Call binaryImage = imclearborder(binaryImage) to get rid of the single line around the perimeter.
- Invert the image: binaryImage = ~binaryImage, so now black circles are white
- Call binaryImage = imclearborder(binaryImage) to get rid of the large (now white) background
- Call bwlabel: [labeledImage, numberOfCircles] = bwlabel(binaryImage). This gives you the count.
- Call measurements = regionprops(labeledImage) if you want other measurements
Counting objects in an image
146 views (last 30 days)
Show older comments
subrajeet
on 7 Jan 2013
Commented: Beenish Ishtiaq
on 5 Aug 2021
Dear Researchers, I have a problem of counting objects in an image. Suppose I have the below image (below link)
which i got after segmentation.
how can I count the small circles in the middle of the image. Kindly suggest
Subrajeet
0 Comments
Accepted Answer
Image Analyst
on 7 Jan 2013
Edited: Image Analyst
on 7 Jan 2013
2 Comments
Image Analyst
on 7 Jan 2013
Edited: Image Analyst
on 7 Jan 2013
It's just basic, elementary image segmentation via intensity thresholding. Nothing fancy at all. See updated instructions above.
More Answers (3)
Thomas
on 7 Jan 2013
Edited: John Kelly
on 8 Nov 2017
You could use techniques shown in many MATLAB image processing demos.
SANAHA PATHAN
on 25 Jul 2016
i want to count number of silkworm eggs from this figure, i used L=Count(binaryfig) this command for counting number of objects ,this command works but i actually dont know what actually this command count,and result of this command is different every time
3 Comments
anna
on 5 Oct 2016
Edited: anna
on 5 Oct 2016
hey I segmented blue nuclei from a blood smear image using k means color segmentation. how can i count the number of segmented blue nuclei. could u please help me with the code for counting [ cant use imfindcircle because the segmented nuclei are not exactly cirlce in shape ]
Image Analyst
on 8 Oct 2016
Beenish Ishtiaq
on 3 Aug 2021
Need code for counting different shapes separatelly which shows each shape result respectively in different edit boxes in GUI matlab..
2 Comments
Image Analyst
on 3 Aug 2021
See my attached demos. I find that usually using findpeaks() to count the number of peaks in the plot of distance of perimeter from centroid works pretty well.
See Also
Products
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!