Matlab's Feature Matching built-in function

Hi,

I am using the Matlab's built-in functions to detect and match features between images. I noticed that the performance of the function "matchFeatures" is much higher than the results found in the literature for same methods such as, ORB, SURF, KAZE, and BRISK.

My question is, what is/are the technique(s) used in feature matching to obtain this performance? Did they use any of the following methods in feature matching: "Bruteforce", "Flann", and/or "RANSAC"?

Regards,

[Rik]

The first step would be to read the documentation. Did you do that?