Real-Time Ensemble-Based Tracker with Kalman Filter

摘要: This work presents an ensemble-based visual object tracker called KFebT. This method can fuse using a Kalman Filter the result of several out-of-the box trackers or specialist methods that solve parts of the problem, like methods that only estimate the target scale variation. Our purpose in joining multiple trackers is to take advantage of the different strengths and weaknesses of each approach. The proposed fusion method is simple and needs no training; it just needs the tracker output result and a confidence measure for the result of …