Histograms represented multi-cue integration based visual tracking

Visual tracking is a hot issue in computer vision, and visual target modeling plays an important role in the robustness and effectiveness of tracking system. In order to adapt target´s observation to the complex tracking environment, this paper proposed a multi-cue integration method, which extracts color, edge, and texture features using histogram method, and integrates them adaptively. By making use of the geometrical and rotation invariance characteristic of histogram, the proposed method shows robustness to variant tracking environments. Experiments on variable videos demonstrate the effectiveness of the proposed method.