Head Tracking Based on Histogram and Shape Model
Qingshan Liu, Songde Ma, Hanqing Lu · 2013
A new model-based method is presented for real-time tracking a person’s head in gray image sequences. Multidimensional receptive field histogram and ellipse shape model is used to describe the tracking object (head). A robust nonparametric technique called mean shift algorithm is adopted to estimate the most probable head location in the current frame, in which histogram matching is used during mean shift iteration. In order to locate the head more accurately and obtain the best scale size of the head, a local search for maximizing the gradient magnitude around the boundary of the elliptical head is preformed after mean shift estimation. It is demonstrated to be a real-time tracker and robustness to scale variation, arbitrary camera movement, partocclusion and so on, for several image sequences. 1.