A visual tracking system includes an image capturing device capturing image frames, a processor receiving the frames to determine tracking objects in the image frames and distances of the tracking objects, and then classify the mark into several object group according to their distances, a mark generating device generating marks relating to the tracking objects respectively, a focusing device having plural of lens members to refract the marks respectively according to the tracking objects they mark. As a result, the mark images on a target will have different image distances to meet the tracking objects.