In this approach, silhouette is fetched from detected object. In this technique, silhouettes are tracked by matching the shape or by evaluation of contour by calculating the object regions in consecutive frames. This method uses the information which is saved inside the object region. This region’s information may be density and shape and models.