OpenCV/C++ - SIFT keypoint detection and corresponding descriptors -


I have an image, I used the edge of the cannie detector, and then I apply it to a lie.

Result at:

Enter image details here

I did not understand the result. Why do not I have some interest points on some black areas and do not expect too much? Also, what is the size of the circle?

I took a look at the internet but it is not completely clear and I have not got any information about it.

?

"Scale-Orient Feature Transform" for SIFT. It detects the main points on many scales, the circles on your image probably refer to the scale at which the main point was detected.

The location of keypoint in black areas does not appear completely in front of me: If you think that the area they see is 2 or 3 times the size of the circle, much more than the main points in the Black Zone It is understandable: it does not detect only a black zone, but for example, such an area; Which is specific: Enter image details here

Important note One main point does not mean that only the exact point is different from the rest; This means that with that point, it is especially with its surroundings. In SIFT case this means that even if you scale the image, SIFT will be able to locate the same point (or zone). If you wish, the "center" of that area is the location of the main place returned.


Comments

Popular posts from this blog

import - Python ImportError: No module named wmi -

Editing Python Class in Shell and SQLAlchemy -

c# - MySQL Parameterized Select Query joining tables issue -