Visual Object Tracking Based on Dynamic Weights and Gaze Density

This study proposes an adaptive algorithm for tracking objects based on the dynamic weights scheme and gaze density. The proposed adaptive algorithm is to amend the deficiency on tracking moving objects that are not driven by inertia. Using the dynamic weights scheme and the information of the gaze point density, possible areas where the objects may appear are constantly researched to target and trace the aimed objects. It is proven to be effective on tracking non-inertia driven moving objects.