The method of positioning moving human based on shape complexity

Positioning moving human in the video sequence of complicated environmental conditions is still a challenging task in computer vision at present, which not only needs to efficiently eliminate varieties of background interferences, but also satisfies real time request in application as well. The shape complexity is applied to position moving human in this paper. Firstly, we adopt the improved frame difference method to extract the moving regions from the video sequence. Secondly, we process the moving regions with the mathematical morphology. Finally, with the shape complexity and the width-height ratio, we can analyze the moving regions and determine whether the regions are moving human body. The experimental results show that the proposed method is feasible in positioning moving human and reduces the phenomenon of omissive detection and error detection efficiently.