Object Detection with Multiple Motion Models

Abstract

Existing joint detection and tracking algorithms generally assume one single motion model for objects of interest However, in real world many objects have more than one motion model In this paper we present a joint detection and tracking algorithm that is able to detect objects with multiple motion models For such an object, a discrete variable is added into the object state to estimate its motion model In this way, the proposed algorithm will not fail to detect objects changing their motion models as the existing algorithms Experimental results show that our proposed algorithm has a better performance than the existing joint detection and tracking algorithms with different single motion models, in detecting objects with multiple motion models.