Circle Detection From Live Feed via Basler Camera but It Detects Wrong Circles and Slow?

Hello everyone,

I am trying to detect circles from live feed via Basler camera but generally it is working wrong and so slow. I am using Hough Circle Detection after create object of my camera with gentl-1 adaptor.

Also my circles has 8cm diameter but this code detects bigger or smaller circles, too.

There is my code;

vid = videoinput('gentl', 1);

set(vid, 'FramesPerTrigger', Inf);

set(vid, 'ReturnedColorspace', 'rgb')

vid.FrameGrabInterval = 25;

start(vid)

while true

e=getsnapshot(vid);

e=imresize(e,[768,1024]);

ks=rgb2gray(e);

[centers, radii]=imfindcircles(ks,[20,60],'ObjectPolarity','dark','Sensitivity',0.9);

imshow(e);

hold on;

viscircles(centers,radii,'EdgeColor','b');

end

What can I add or do to the code I wrote to make the system work more accurately and quickly?

[Image Analyst]

You forgot to attach an image. So . . . buy a faster computer? Or use a different algorithm (which we can't suggest because you forgot to attach the image).