Help on detecting a a curve form

asked 2012-11-09 08:20:39 -0600

peterp
11 ●1

Hello Gurus,

I am hoping that the wise would guide the misguided here :).

I am trying to detect a curve of a certain shape and its position from a signal as shown below: image description . Due to noise and other variations, the curve may not be exact so I can not use simple number matching. I was wondering if there is something in OpenCV which I can use for this. Note that I will need to detect curves of different shapes and their position in the signal but if I know to detect one type, I can use the same method to detect other types.

Your help is appreciated. Thanks, Peter

edit retag flag offensive close merge delete

Comments

do you get an image like in your posting or do you get a one-dimensional float array or similar?

Ben ( 2012-11-09 08:43:58 -0600 )edit

I would get a 1D float array. Thanks.

peterp ( 2012-11-09 08:58:07 -0600 )edit

add a comment

1

answered 2012-11-09 11:56:53 -0600

Ekalic
83 ●1 ●5 http://www.tldvision.com/

This question is more related to signal processing than to computer vision since the input in 1D. Anyway, there is a large number of potential methods that you can use. For practical reason, I would start by obvious choices (detection of maximum in signal). If the noise in the data does not allow to use it, you should think how to model the noise. This can be done either by encoding it manually or learning the expected variations from training data. What kind of signal is it by the way?