支持多种文字的视频字幕叠加工具设计  

Design of a multi-language video subtitle overlay tool

作　　者：王亚[1] 褚晶辉[1] 刘子玉[1] 吕卫[1] 

机构地区：[1]天津大学电子信息工程学院,天津300072

出　　处：《信息技术》2015年第9期118-120,共3页Information Technology

摘　　要：文中设计了一种基于开源库OpenCV、FreeType和VLC的多种文字视频字幕叠加工具。该工具首先读取字幕文件中的时间信息和文本信息,然后按时间信息解码对应的视频帧,最后将字幕的文本内容绘制到视频帧中。克服了OpenCV只支持西文字符绘制、不支持中文字符的缺点,并支持多种字体的绘制。该工具可在视频截图上批量叠加不同字体的字幕,能够很便捷地生成大量字幕素材,用于视频字幕检测算法的设计或评估。This paper presents the design of open source libraries OpenCV, FreeType and a multi-language video subtitle overlay tool based on the VLC. The tool reads time information and text information from the subtitle files at first, and then decodes video frames according to the time information. Finally, the text of the subtitle is drawn on the extracted frames. The tool overcomes the shortcoming of OpenCV which only supports western characters and does not support Chinese characters, and is capable of drawing multiple fonts. The tool can overlay subtitle in different fonts to the video snapshot in a batch mode, which is convenient to provide massive subtitle material for designing or evaluating video subtitle detection algorithms.

关 键 词：字幕检测 OPENCV FREETYPE VLC 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]