Ask Your Question

What is the sampling method used by cv2.SIFT.detect?

asked 2015-12-01 19:19:54 -0500


I'm new to OpenCV and SIFT algorithm. I'm using the python version of the library. I know that I can create lots of KeyPoint objects manually and send them to the cv2.SIFT.compute method to get their SIFT descriptors. Although, OpenCV offers another method (cv2.SIFT.detect) to automatically detect a set of KeyPoint objects, given an image.

My question is: what is the sampling method adopted by the cv2.SIFT.detect method?

Additionally: where in the literature can I find a detailed description of such method?

Best regards!

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2015-12-02 01:09:30 -0500

LorenaGdL gravatar image

I guess this tutorial will help with a high-level description of the algorithm: Also mentioned there the original paper: David G. Lowe. 2004. Distinctive Image Features from Scale-Invariant Keypoints. Int. J. Comput. Vision 60, 2 (November 2004), 91-110.

edit flag offensive delete link more



Also a side warning: SIFT is patent-protected in some countries. So, be careful if you plan to use it in commercial applications

LorenaGdL gravatar imageLorenaGdL ( 2015-12-02 01:10:51 -0500 )edit

Question Tools

1 follower


Asked: 2015-12-01 19:19:54 -0500

Seen: 2,093 times

Last updated: Dec 02 '15