视频中的文字信息为视频语义的理解提供了重要信息,该文提出一种改进的视频标题文字检测方法,该方法不仅能检测视频中文字出现位置,而且能检测到标题文字出现的时间边界.对数小时的视频标题检测实验表明,该方法是有效的,总有效率可达到80%左右.
Textual information brings important semantic clues to video content analysis. This paper proposes an improved method for video caption detection and extraction. The method not only detects the location of caption but also determines the temporal boundaries. Experimental results on several hours video show the efficiency of the method, which achieves the accuracy of about 80%.