Identify irregular forms.
Hi this is an image of a wheat apex. What im trying to do is count the amount of "fingers" located on the sides of the apex. However I'm a little bit lost deciding which function I shall use because of the irregularity of this fingers/blobs. Is it even posible to count them? I have a couple of ideas in mind, but I don't know how to translate them into functions in OpenCv.
"I have a couple of ideas in mind" - definitely append those !
Hi Berak!. Well, I was thinking if I could draw a tangent line in each "finger tip" then I would be able to count each line and see the result. Or maybe identify each tip as a semi-circle, but I don't know if such and irregular shape could be consider a semi-circle. Maybe identifying the gaps between the bumps is easier. That's all I got. However I don't know if any of those things can be done in OpenCv. Thank you very much for your reply.
A challenging task! The left part could be done by efficient edge detection, but the contrast on the right kind of ruins that. Is it a symmetric shape or not?
It's not symmetric, and I will be uploading a better shot with a better contrast. So "efficient edge detection" tell me more about that haha :D. I know about "Canny Edge Detection", but I don't know how to analyse the final shape that canny gives you in order to count the bumps. Thank you very much StevenPuttemans!
Hmm the efficient detections is indeed the problem. What about going for a Sobel operator in Y direction, looking for a decent threshold response, then trying to fit houghlines of a fixed length range? Then you could analyze the relation between several found lines, coupling two lines as 1 segment?
I think you should look up some interesting literature on this topic. OpenCV won't contain a ready to use piece of software for this: