首页 | 官方网站   微博 | 高级检索  
     

基于音频流的电视智能监测系统设计
引用本文:崔朝阳,刘晓星,韩疆,颜永红.基于音频流的电视智能监测系统设计[J].微计算机应用,2005,26(6):687-690.
作者姓名:崔朝阳  刘晓星  韩疆  颜永红
作者单位:1. 广电总局五七三台,北京102209
2. 中科院声学所,北京,100080
摘    要:广电总局的电视监测业务已经实现了设备控制自动化和卫星信号采集的数字化、信息化和网络化,但基于内容的异态事件监测和信息处理还是完全依赖人工完成.语音处理、语音识别和关联检索等技术的发展,为电视监测业务智能化提供了可能.本文介绍了电视监测业务智能辅助系统的架构,该系统能够自动定位电视节目,把电视新闻语音转化为文字,对敏感语言内容预警,并关联聚类相关信息,方便人工后续处理.

关 键 词:语音识别  说话人识别  字幕识别  固定音频检索  基于内容的多媒体检索  关联聚类  电视监测业务
收稿时间:2005-04-07
修稿时间:2005-04-07

The Design of Intelligent Television Monitoring System
CUI Chaoyang,LIU Xiaoxing,HAN Jiang,YAN Yonghong.The Design of Intelligent Television Monitoring System[J].Microcomputer Applications,2005,26(6):687-690.
Authors:CUI Chaoyang  LIU Xiaoxing  HAN Jiang  YAN Yonghong
Affiliation:1.The State Administration of Radio Film and Television; 2.Institnte of Acoustics, Chinese Academy of Science, Beijing 100080
Abstract:Broadcasted TV program content monitoring system has achieved automatic device control and satellite signal capturing by utilizing information technology. But unexpected content/event monitoring and information processing still rely upon human. With the advancement of speech processing, speech recognition and information clustering/ retrieval technology, it is now possible to monitor TV content automatically. This paper presents the supporting monitoring system for TV program. This system can automatically locate specified TV content, convent the audio of TV news into written language, alert upon preset sensitive content, and use automatic clustering technology to facilitated subsequent processing.
Keywords:speech recognition  speaker recognition  caption optical character recognition  content based multimedia retrieval  associating and clustering  television monitoring operation
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号