Probabilistic Cue Integration for Real-Time Object Pose Tracking

Robust real time object pose tracking is an essential component for robotic applications as well as for the growing field of augmented reality. Currently available systems are typically either optimized for textured objects or for uniformly colored objects. The proposed approach combines complementary interest points in a common tracking framework which allows to handle a broad variety of objects regardless of their appearance and shape. A thorough evaluation of state of the art interest points shows that a multi scale FAST detector in combination with our own image descriptor outperforms all other combinations. Additionally, we show that a combination of complementary features improves the tracking performance slightly further.