ALVT Adaptive model update via fusing peak-to-sidelobe ratio and mean frame difference for visual tracking (Acta Photonica Sinica | Acta Phot Sin)