问题
I would like to ask how is SIFT algorithm done in C++.
I understand that SIFT extracts keypoints. In that case, what will be the value represented by the keypoints extracted by SIFT algorithm? And how can I compare and detect similar keypoints?
I will be using SIFT in my thesis. Can someone please discuss this to me in a more simpler manner? I'm a good programmer but I can't seem to understand how to implement SIFT. I haven't had a class on Image processing.
回答1:
I suggest you start from the original article by Lowe to understand the terms and how they're calculated. You can find this and other articles at Google Scholar.
You can probably find implementations of SIFT in both C++ and Matlab, unless you must implement it yourself.
The bottom line is that, although the algorithm is pretty simple, it's only simple if you understand the underlying image processing algorithms. You'll have to beef up on them to understand SIFT.
回答2:
OpenCV library has implementation of SIFT and some algorithms to find matches.
Here is an online documentation for features2d OpenCV module.
回答3:
Use this tutorial. It explains SIFT in a very user-friendly manner.
来源：https://stackoverflow.com/questions/6960455/sift-and-keypoint