作者: Zhaohui Sun , Majid Rabbani , Gabriel Fielding
DOI:
关键词:
摘要: A method for video filtering of an input sequence by utilizing joint motion and noise estimation comprises the steps of: (a) generating a motion-compensated from plurality estimated fields; (b) spatiotemporally compensated sequence, thereby producing filtered, sequence; (c) estimating standard deviation difference between deviation; (d) scale factor (e) iterating through to using previously obtained step generate in perform until value level approaches unknown whereby is then characterized finally determined deviation.