snavely / vocabtree2 Goto Github PK

Noah,

VocabTree2 works very well for me with SIFT descriptors, as it was designed. I am looking for a solution using binary descriptors, such as BRISK. From what I've been reading, it appears that VocabTree2 may not generalize well to binary descriptors since it uses k-means and Euclidean distance, while for binary descriptors the Hamming distance and other ways of clustering may be more appropriate. I wonder if you have any insights on this, and if by chance any pointers to relevant software. Thank you.

I am using a gpu implementation of SURF to get descriptors and it gives me descriptor values in float [-1 1]. If I scale that to [0 255] I lose a lot of precision. Is there a way I can train the Vocabulary tree on float descriptor values instead of integer values?

Thanks a lot!

Hi snavely, I use your code on Microsoft Visual Studio2013, and there are some link error after I modify somewhere to guarantee the compile pass, the error remind that there are lacking of some lib flie, and I find there are only one lib named "zdll.lib", so I am confused that whether the code could run on ubuntu only or I do a wrong opration?
Looking foward your reply!

One minor thing I noticed is that the branching factor can be tricky to set. The recommended value of 500000 in my experience results in an image matching well to itself but not to its neighbors from a video sequence. Perhaps some heuristics could be used here, maybe the square root of number of features?

I was wondering if there's any support or a quick hack to add images to an already created vocab tree.

Thanks a lot!

Hi snavely, there are some link error remind me that there are missing some lib file, I write the K-means replace your Kmeans.cpp, and build the vocabulary tree successfuly. But when I run the VocaBuildDB.cpp, the link error appear again, and I find that the function "BuildANNTree" just define but not implement, so would you mind tell me what the function do or how can I deal with the problem?
Thanks for your reply!