I want to use N-SIFT (n-dimensional scale invariant feature transform)feature points to detect some spatio-temporal scale invariant regions in a video. I have got the code from the following link http://www.insight-journal.org/browse/publication/207
It builds the tool on Cygwin platform and uses ITK. No detailed description is given on how to build it. So if any one using this tool/library please help me to build it. Or please mention any other library for the same if available.