Objects tracking with adaptive correlation filters and Kalman filtering
Sergio E. Ontiveros-Gallardo, Vitaly Kober · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2015
Object tracking is commonly used for applications such as video surveillance, motion based recognition, and vehicle navigation. In this work, a tracking system using adaptive correlation filters and robust Kalman prediction of target locations is proposed. Tracking is performed by means of multiple object detections in reduced frame areas. A bank of filters is designed from multiple views of a target using synthetic discriminant functions. An adaptive approach is used to improve discrimination capability of the synthesized filters adapting them to multiple types of backgrounds. With the help of computer simulation, the performance of the proposed algorithm is evaluated in terms of detection efficiency and accuracy of object tracking.