Ask Your Question
2

Is FREAK scale and rotation invariant?

asked 2012-08-03 21:22:44 -0600

this post is marked as community wiki

This post is a wiki. Anyone with karma >50 is welcome to improve it.

I've read news about FREAK in OpenCV 2.4.2. Is FREAK scale and rotation invariant, like SURF? Is it as good as SURF?

edit retag flag offensive close merge delete

Comments

1

Goodness is a relative measure - you should try both of them for your specific situation, and choose the best trade-off between quality, speed, and probably other factors.

sammy gravatar imagesammy ( 2012-08-06 03:15:48 -0600 )edit

3 answers

Sort by ยป oldest newest most voted
4

answered 2012-09-18 05:06:35 -0600

this post is marked as community wiki

This post is a wiki. Anyone with karma >50 is welcome to improve it.

SURF is a detector like FAST. FREAK is a descriptor extractor. It has more sense to compare it with SIFT.

Actually, in the paper they say they use FREAK descriptors together with a pyramidal FAST detector that is called AGAST.

So answering your question:

  • It is rotation and scale invariant.

  • It is supposed to be a more robust descriptor than SIFT. I am going to test it but still don't have results.

edit flag offensive delete link more

Comments

2

SURF isn't just a detector (which uses the fast-hessian method), it also is a descriptor. You can extract 64 or 128 dimensional descriptors from your detected keypoints.

Ben gravatar imageBen ( 2012-11-05 12:15:13 -0600 )edit
2

answered 2012-08-05 21:05:19 -0600

this post is marked as community wiki

This post is a wiki. Anyone with karma >50 is welcome to improve it.

Hi!

I recommend you to read the paper FREAK: Fast Retina Keypoint. Alahi et al. CVPR2012.

The authors claim that FREAK is scale and rotation invariant and even more robust than SURF.

I hope this helps.

edit flag offensive delete link more
2

answered 2012-11-05 11:26:09 -0600

this post is marked as community wiki

This post is a wiki. Anyone with karma >50 is welcome to improve it.

It's actually a little bit more subtle: the pairs used by FREAK were chosen for their matching performance after a scale-sensitive detection of keypoints (the KP detector used is the same as BRISK, i.e. a multi-scale FAST + orientation detection). Hence, FREAK is rotation and scale invariant only up to the sensitivity of this detector and is not per se invariant.

edit flag offensive delete link more

Question Tools

1 follower

Stats

Asked: 2012-08-03 21:22:44 -0600

Seen: 2,644 times

Last updated: Nov 05 '12