This paper presents a framework to automatically replant timestamps for panorama video surveillance. The system mainly comprises six modules: 1) timestamp localization; 2) image registration; 3) timestamp pixel recovery; 4) color correction; 5) timestamp recognition; and 6) panorama video generation. The proposed framework first develops a method to localize the timestamp by employing the method of locating digital video clocks. Then, a fast and novel method is developed to recover the pixels covered by timestamps. The proposed method has fast speed because a fast image translation estimation ...
