网络视频内容的识别和过滤综述  

Survey on recognition and filtering of network video content

作　　者：彭乐[1] 薛一波[2] 王春露[1] 

机构地区：[1]北京邮电大学计算机科学与技术学院,北京100876 [2]清华大学信息技术研究院,北京100084

出　　处：《计算机工程与设计》2008年第10期2587-2590,2634,共5页Computer Engineering and Design

摘　　要：随着网络视频的迅猛发展和广泛使用,网络不良视频的识别和过滤日益重要。通过对图像内容识别与过滤、视频结构分析与检索两个领域技术发展的分析,阐述了一种综合利用视频时域分割、关键帧提取、图像内容识别及皮肤检测等视频分析方面关键技术的解决方法。该方法简单,并且容易实现。此外,介绍了目前网络视频内容识别和过滤的研究现状和主要应用,分析了其面对的主要问题及未来发展趋势。With the rapid development and extensive use of network video, it is important for recognizing and filtering objectionable video. By analyzing the development ofrelated fields, such as image content recognition and filtering, and content-based video analysis and retrieval, one solution is proposed in works which is based on the key technologies of video temporal segmentation, key frame ex- traction, image content recognition, and skin detection. Otherwise it is simple to understand and easy to be implemented. Then, the current research status and the main applications of the recognition and filtering of video content of network are presented. Finally, its main challenges and future trends are analyzed and given.

关 键 词：时域分割 关键帧提取 图像内容识别 皮肤检测 视频内容识别和过滤 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]