摘要
Authors present a novel object tracking based video parsing algorithm for the first step to video analysis. Segment the frame of video clip to find the biggest block. After analyzing the mean color and area of the block’s, track it and generate the boundary information of the video shot’s. This algorithm could be used in real time video segmentation and shows a good performance in the experiment.
Authors present a novel object tracking based video parsing algorithm for the first step to video analysis. Segment the frame of video clip to find the biggest block. After analyzing the mean color and area of the block's, track it and generate the boundary information of the video shot's. This algorithm could be used in real time video segmentation and shows a good performance in the experiment.
出处
《四川大学学报(自然科学版)》
CAS
CSCD
北大核心
2005年第2期420-423,共4页
Journal of Sichuan University(Natural Science Edition)