A method for decision fusion of target trackers running on different band image sequences

In this work, an Euclid distance based decision fusion framework for the trackers running on the feature sequences extracted from different band images is proposed. In the experiments, four correlation based scale invariant tracker are run on intensity and edge image sequences extracted from the visual and infrared bands. The target locations obtained from each tracker are fused using the proposed method. The experimental studies show that, the proposed tracker decision fusion scheme makes the tracker more robust to track loss conditions.