Human motion tracking in natural scene based on adaptive model

Human motion real-time tracking in nature scene is a hot issue in the field of computer vision. Because of the complex of the natural scene, it is difficult to strengthen the robustness of the algorithm. Aiming at the problem, a background model that can be adaptively refreshed, is brought forward in this paper. Using the statistical model the moving body can be segmented, and a real-time adaptive Kalman filtering tracking model is applied to the moving body tracking. Experimental results in nature scene prove the proposed validity of the algorithm.