联合时空SIFT特征的同源视频检测

通过对视频帧序列时空特性的分析，采用“局部趋同，全局异化”的策略，提出了一种联合时域和空域SIFT点特征的特征提取方法。实验表明，基于该特征的同源视频检测方法对于一定的视频变化具有较好的鲁棒性和检测精度。

Saved in:
Bibliographic Details
Published in电子技术应用 Vol. 38; no. 3; pp. 130 - 133
Main Author 张瑞年 于洪涛 李邵梅
Format Journal Article
LanguageChinese
Published 国家数字交换系统工程技术研究中心,河南郑州,450002 2012
Subjects
同源视频检测
联合时空SIFT特征
视频帧序列时空特性
联合时空SIFT特征
视频帧序列时空特性
同源视频检测
Online AccessGet full text
ISSN0258-7998
DOI10.3969/j.issn.0258-7998.2012.03.042

Cover

More Information
Summary:通过对视频帧序列时空特性的分析，采用“局部趋同，全局异化”的策略，提出了一种联合时域和空域SIFT点特征的特征提取方法。实验表明，基于该特征的同源视频检测方法对于一定的视频变化具有较好的鲁棒性和检测精度。
Bibliography:detection of coderivative video; spatiotemporal characteristics of video frames sequences ;spatiotemporal SIFT char-acteristics
11-2305/TN
Video features extraction is the important part of detection of coderivative video .SIFT features are local features with excellent performance. By anglicizing the temporal and spatial characteristics of video frames sequences SIFT feature, this paper takes the ＂local convergence, global alienation＂ strategy, and proposes a feature extraction method of spatiotemporal SIFY characteristics. Experiments of detection of coderivative video based on the method show that the proposed co-features has better robustness against certain changes of video sequences ,and also has a better detection accuracy than some existing methods.
Zhang Ruinian, Yu Hongtao, Li Shaomei （National Digital Switching System Engineering ＆ Technological R＆D Center, Zhengzhou 450002, China）
ISSN:0258-7998
DOI:10.3969/j.issn.0258-7998.2012.03.042