作者: Ovadya Menadeva , Itzhak Wilf , Hayit Greenspan
DOI:
关键词:
摘要: A method of selecting key-frames (230) from a video sequence (210, 215) by comparing each frame in the with respect to its preceding and subsequent for redundancy where comparison involves region motion analysis. The is optionally pre-processed detect graphic overlay. key-frame set post-processed (250) optimize resulting face or other object recognition.