Evolutionary Fuzzy Adaptive Motion Models for User Tracking in Augmented Reality Applications

In Augmented Reality (AR) applications, tracking the movements of user is the one of the most crucial issues. Because of the unpredictable structure of human movement, tracking the user with classical robot tracking methods can cause inaccurate result. In this study, motion different models for increasing the precision of human tracking using GPS-INS receiver was developed. First, a fuzzy motion model was developed and this model was improved using an evolutionary algorithm. With these algorithms allowing to choose between different motion models, transition among the motion models was achieved in real time and precision was increased for human tracking.