提出了整帧信噪比分段加权法,在低信噪比的情况下,提高系统的识别率。
Proposing a segment weight method of the whole SNR which can improves the identification ratio under the lower SNR circumstance.
其基本思想是先对新闻视频进行镜头边界检测;其次,通过文字检测算法检测包含有主题字幕文本的帧来获得新闻故事分段的线索;
The basic idea is to detect the shot boundaries first for news video, and then frames containing topic caption texts are identified to get news story segmentation cues using text detection algorithm.
应用推荐