摘要
介绍文化遗产视频检索系统的设计与实现方法。描述采用MVC架构的系统的总体结构及其主要功能模块。针对网络带宽窄、视频传输数据量大的特点,提出一种新的支持基于内容的视频检索算法——基于"视频指纹"的检索方法,采用RGB颜色空间平均值作为视频特征。设计了相应的匹配检索算法,实验结果表明该算法具有较高的检索准确度,查全率比原算法提高了33.33%。
The paper presents the design and implementation of cultural heritage video retrieval system. The system structure with MVC pattern and the function of each module are described. Aiming at the narrow network bandwidth and mass data transferring, a new Content-based Video Retrieval（CBVR） method using video fingerprints is proposed. In the algorithm, the average value of RGB color channels is calculated as the characteristics of video sequences, and the corresponding video retrieval algorithm is developed. Experimental results show that the algorithm gets better performance than original one, the nicety has been improved by 33.33%.
出处
《计算机工程》
CAS
CSCD
北大核心
2008年第10期226-228,共3页
Computer Engineering
基金
国家自然科学基金资助项目(60672050)
北京市自然科学基金资助项目(4062005)
北京市属市管高等学校人才强教计划基金资助项目(00627)
关键词
MVC架构
基于内容的视频检索
视频指纹
MVC pattern
Content-based Video Retrieval（CBVR）
video fingerprints
