AForge.NET

  :: AForge.NET Framework :: Articles :: Forums ::

Circle Detection

The forum is to discuss topics related to different areas of image processing and computer vision.

Circle Detection

Postby Pavinkumar » Fri Aug 17, 2018 11:01 am

Hi,

How to detect a circular shape in an binary image when the circle is partially covered / disturbed by some other blobs. Erosion / Dilatation might help a bit, but, I don't want to compromise the full circular area that's been covered in the raw image.

Please check the attached images for reference.

Any help on this would be appreciated !

Thanks
Attachments
Raw_Image_3.PNG
Raw_Image_3.PNG (5.56 KiB) Viewed 552 times
Raw_Image_2.PNG
Raw_Image_2.PNG (5.96 KiB) Viewed 552 times
Raw_Image_1.PNG
Raw_Image_1.PNG (2.92 KiB) Viewed 552 times
Pavinkumar
 
Posts: 2
Joined: Fri Aug 17, 2018 10:46 am

Re: Circle Detection

Postby andrew.kirillov » Mon Aug 20, 2018 6:57 am

Hello,

One thing to try is edge detection and then Hough Circle Transformation. But you need to know radius of the circle you want to detect.
With best regards,
Andrew


Interested in supporting AForge.NET Framework?
User avatar
andrew.kirillov
Site Admin, AForge.NET Developer
 
Posts: 3325
Joined: Fri Jan 23, 2009 9:12 am
Location: UK

Re: Circle Detection

Postby Pavinkumar » Thu Aug 23, 2018 2:09 pm

Andrew,

Thanks for the suggestion. It did work out for me. I did Canny Edge Detection & then Hough Circle Transform.

The exact radius of the circle is unknown for me. But, I know an approximate range [ex : 20 to 60]. I did the same procedure for all the possible values in the range & then filtered out the one with maximum Intensity. This did work out for me to locate the circle i am looking for. This even helped me filter out the biggest circle if there are multiple circles in the given image.

Thank you again :)
Pavinkumar
 
Posts: 2
Joined: Fri Aug 17, 2018 10:46 am




Return to Image Processing and Computer Vision