[1]郭智强,黄志华.基于类间距离和准则的镜头分割[J].南京师范大学学报(工程技术版),2008,08(04):108-111.
 Guo Zhiqiang,Huang Zhihua.Shot Detection Base on Maximum-Between-Cluster-Distance[J].Journal of Nanjing Normal University(Engineering and Technology),2008,08(04):108-111.
点击复制

基于类间距离和准则的镜头分割
分享到:

南京师范大学学报(工程技术版)[ISSN:1006-6977/CN:61-1281/TN]

卷:
08卷
期数:
2008年04期
页码:
108-111
栏目:
出版日期:
2008-12-30

文章信息/Info

Title:
Shot Detection Base on Maximum-Between-Cluster-Distance
作者:
郭智强;黄志华;
福州大学数学与计算机科学学院, 福建福州350002
Author(s):
Guo ZhiqiangHuang Zhihua
College of Mathematics and Computer Science,Fuzhou University,Fuzhou 350002,China
关键词:
视频检索 镜头分割 类间距离
Keywords:
v ideo de tection shot boundary detection class distance
分类号:
TP391.41
摘要:
镜头切换检测是视频检索研究领域中一项基本的、重要的技术.然而传统的镜头边界检测方法都只考虑了视频相邻两帧的信息,没有充分考虑到视频序列中前后的连续性,从而容易导致误检测.基于此,提出了一种基于类间距离和准则对镜头边界进行检测的新方法,该方法能够充分利用视频序列中的连续性信息,克服传统镜头边界检测方法的不足,有效地提高了检出率及精确率.大量实验表明,该方法是行之有效的.
Abstract:
The shot chang e detection is a basic and im portant techno logy in v ideo retr ieva.l But the trad itiona l sho t boundary detec tion m ethods only conside r the two adjacent v ideo inform ation, and do no t take fu lly in to account the prev ious and consequent v ideo sequences in the continuity and thus easily lead to error detection. Therefore a new sho t boundary detection a lgor ithm based on max imum-betw een-c luster-d istance is proposed. Them ethod can m ake full use o f v ideo sequences in the continuity o f inform ation, and overcom e the traditiona l boundary- less than the detection m ethod. La rge num ber o f exper im ents show s tha t the algor ithm has e ffective ly improved the de tection rate and prec is ion.

参考文献/References:

[ 1] Ngo C W, Pong T C, Ch in R T. V ideo partition ing through temporal slices analysis[ J]. IEEE T ransactions on C ircuits and
System s for V ideo Technology, 2001, 11( 8): 941-953.
[ 2] 林通, 张宏江, 封举富, 等. 镜头内容分析及其在视频检索中的应用[ J]. 软件学报, 2002, 13( 8): 1 577-1 585.
Lin Tong, Zhang H ong jiang, Feng Jufu, et a.l Sho t content analysis fo r video re tr ieva l applications[ J]. Journal o f Softw are,2002, 13( 8): 1 577-1 585. ( in Ch inese)
[ 3] Boon-Lock Yeo, Bede L iu. Rap id scene analysis on com pressed video[ J]. IEEE Transactions on C ircuits and Systemsf or V ideo Technology, 1995, 5( 6) : 533-544.
[ 4] Idr is F, Panchanathan S. Rev iew o f image and v ideo index ing techniques[ J]. Journa l o fV isua l Comm un ica tion and Image Rep
resenta tion, 1997, 8( 2): 146-166.
[ 5] Courtney J D. Autom atic v ideo index ing v ia object m otion ana lys is[ J]. Pa ttern Recogn ition, 1997, 30( 4): 607-625.
[ 6] Yasuyuk iNaka jima. A v ideo brow sing using fast scene cu t detec tion for an e fficien t netw orked v ideo da tabase access[ J]. IEICE
Trans on Inform ation and Systems, 1994, E77- D( 9) , 1 355-1 364.for v ideo database. In: Proc ICASSP 96, A ta lanta, GA, 1996.
[ 7] Chang CW, Lee S Y. V ideo content representation, index ing, andm a tch ing in video inform ation system [ J]. Journal o fV isual Commun ica tion and Im age Representa tion. 1997, 8( 2): 107-120.
[ 8] 边肇祺, 张学工. 模式识别[M ]. 2版. 北京: 清华大学出版社, 2000.
B ian Z Q, Zhang X G. Pa ttern Recogn ition[M ]. 2nd ed. B eijing: T singhua University Press, 2000. ( in Ch inese)
[ 9] 丁洪丽, 陈怀新. 基于累积直方图的视频镜头边界检测方法[ J]. 电讯技术, 2008, 48( 3): 65-69.
DingH ongl,i Chen H ua ix in. V ideo shot bounda ry de tection m ethod based on cumu lative histog ram [ J]. Te lecomm un ication Eng ineer ing, 2008, 48( 3): 65-69. ( in Ch inese)

备注/Memo

备注/Memo:
通讯联系人: 黄志华, 副教授, 研究方向: 数据库系统. E-m ail: hzh@ fzu. edu. cn
更新日期/Last Update: 2013-04-24