By nerith
on Tue, 10/23/2012 - 14:46
Hi,
Does anybody knows some details about the feature detection algorithm ?
I would like to create the most optimized tracker for small distances, how does the size of the uploaded image impact the recognition ?
What is the threshold on the number of feature needed to be recognized in real time ? IE, if my trackable include 100 featured points, how many should be detected on runtime before the trackable is detected ?
Does the number of featured points on my trackable impact the recognition speed ?
Does the number of different possible trackable impact the recognition speed ?