Old Article - A comparative analysis of SIFT, SURF, KAZE, AKAZE, ORB, and BRISK


I’m trying to understand some of the options available for feature extraction / matching I found the following document comparing many of them (In matlab using OpenCV), wich apply to OpenSFM that use OpenCV also; but no detail in the hahog algorithm.

This was all ready feature by Saijin_Naib but is not referenced in the learning Area.


Thanks for posting this up!

It looks like we’d need to do a bit of work upstream with OpenSFM to tune and adjust the other feature types to bring them up to the flexibility and power of our current SIFT implementation.

Yes, I’ve similarly found not terribly much detail regarding HAHOG in the literature.

Apparently HAHOG is the mixture of two methods of computer vision:

  • Hessian Affine feature point detector
  • HOG descriptor

Here some literature:

Hessian Affine feature point detector

HOG Feature Descriptor found this: