Hello everyone,
I have some trouble to understand SIFT local feature
Could you help me to answer to this question ? just to understand more this approach
how D.lowe fix the number of octaves ?
How they calculate the sigma and K ?
when they extract local candidate key point. Are they passed by all the pixels in the images?
in addition after a DoG we get a set of images how they merge them to have one image where all candidate key points are on it ?
why exactly they use the pyramid, scale factor, sigma and K the DoG ?
what is there effect ?
this is my first part of question
Thank you