稠密子图发现的视频语义挖掘方法

计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (33): 13-14.

稠密子图发现的视频语义挖掘方法

张师林1，2，李和平2，张树武2

1.北方工业大学计算机基础教研室，北京 100144
2.中国科学院自动化研究所数字媒体研究中心，北京 100191

收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-11-21 发布日期:2011-11-21

Dense sub graph based video semantic mining

ZHANG Shilin1，2，LI Heping2，ZHANG Shuwu2

1.Computer Faculty，North China University of Technology，Beijing 100144，China
2.Digital Media Center，Institute of Automation，Chinese Academy of Sciences，Beijing 100191，China

Received:1900-01-01 Revised:1900-01-01 Online:2011-11-21 Published:2011-11-21

摘要/Abstract

摘要： 目前基于内容的视频语义挖掘方法并未考虑到视频的多模态特性，不能够实现对于目前海量涌现视频的自动分析处理任务。针对此问题，提出了基于稠密子图发现的视频语义挖掘方法。该方法对待处理的视频进行中文连续语音识别、视频目标识别和视频文字识别，对于识别结果进行中文分词和词性标注，保留名词和动词作为图模型的顶点，顶点之间的边权重设置为两个顶点所代表的词语的中文语义距离，根据稠密子图发现算法挖掘视频的语义信息。实验结果表明这种方法是有效的。

关键词: 稠密子图, 中文连续语音识别, 视频目标识别

Abstract: Content-based video semantic mining does not take the video multi-modal attribute into account，and the mapping is not in line with the perception of video by mankind.The method above can not carry out the automatic video analysis task.Dense sub graph based semantic mining method is presented.This method integrates Chinese continuous speech recognition，video object recognition and video text recognition.Graph is used to present the video by denoting the words with vertex and the words’ relation with edges.The experimental results show that this method is effective.

Key words: dense sub graph, Chinese continuous speech recognition, video object recognition

张师林1，2，李和平2，张树武2. 稠密子图发现的视频语义挖掘方法[J]. 计算机工程与应用, 2011, 47(33): 13-14.

ZHANG Shilin1，2，LI Heping2，ZHANG Shuwu2. Dense sub graph based video semantic mining[J]. Computer Engineering and Applications, 2011, 47(33): 13-14.