基于优化的DTW算法的人体运动数据检索

摘要
图/表
参考文献
相关文章 (7)

全文: PDF (616 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要随着大量三维人体运动数据库的建立，使得在数据库中实现基于内容的三维人体运动检索面临着诸多困难，文中提出一种分阶段的动态时间变形(DTW)优化算法的人体运动数据检索技术，可有效检索出逻辑上相似的运动。该算法首先对齐两个运动序列的坐标位置，基于窗口距离构造距离矩阵。其次采用基于全局和局部约束的DTW优化算法进行相似度匹配，得到两个运动间的对应关系。最后通过归一化相似度和DTW平均距离分阶段判断运动的相似性。实验结果表明，分阶段的DTW优化算法在提高效率的同时对长度不等的运动能取得较好的检索结果。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	刘贤梅
	赵丹
	郝爱民

关键词 ：三维人体运动, 运动检索, 动态时间变形(DTW)算法

Abstract：With the emergence of many large-scale three-dimensional human motion databases, content-based retrieval of 3D human motion faces many difficulties. A human motion data retrieval technology based on improved dynamic time warping optimization algorithm is proposed, by which logically similar motions can be found effectively. Firstly, the coordinates of two motion sequences are aligned and a distance matrix based on window of frames is constructed. Then, using an optimization algorithm based on global and local constraints, a similarity matching is processed to describe corresponding relationship between two motions. Finally, similar motions are retrieved by two-phase approach with normalization similarity and DTW average distance. The experimental results show that by two-phase DTW optimization approach, better retrieval results for motions which are not aligned in time axis are obtained and the efficiency is improved.

Key words： 3D Human Motion Motion Retrieval Dynamic Time Warping (DTW) Algorithm

收稿日期: 2011-03-25

ZTFLH:

TP391.9

基金资助:黑龙江省科技计划项目(No.GZ09A118)、北京市教育委员会科技计划项目(No.KM200910005020)资助

作者简介: 刘贤梅，女，1968年生，教授，博士研究生，主要研究方向为计算机图形学、人体动画等。E-mail:liuxianmei@yeah。net。赵丹，女，1985年生，硕士研究生，主要研究方向为人体动画。郝爱民，男，1968年生，教授，博士生导师，主要研究方向为计算机图形学、三维逼真、实时绘制等。

引用本文:

刘贤梅，赵丹，郝爱民. 基于优化的DTW算法的人体运动数据检索[J]. 模式识别与人工智能, 2012, 25(2): 352-360. LIU Xian-Mei, ZHAO Dan, HAO Ai-Min. Human Motion Data Retrieval Based on Dynamic Time Warping Optimization Algorithm. , 2012, 25(2): 352-360.

链接本文:

http://manu46.magtech.com.cn/Jweb_prai/CN/ 或 http://manu46.magtech.com.cn/Jweb_prai/CN/Y2012/V25/I2/352

[1] Yu Jun,Zhuang Yueting.A Survey on Visual Media Fusion in Character Animation.Journal of Computer-Aided Design Computer Graphics,2008,20(5): 545-553(in Chinese)
(俞俊,庄越挺.计算机角色动画制作中的可视媒体融合综述.计算机辅助设计与图形学学报,2008,20(5): 545-553)
[2] Liu Xianmei,Zhao Dan,Zhao Ya.Survey on Keyframe Technology of Data-Driven Human Animation.Computer Engineering and Design,2011,32(3): 1006-1009(in Chinese)
(刘贤梅,赵丹,赵娅.数据驱动人体动画的关键帧技术研究综述.计算机工程与设计,2011,32(3): 1006-1009)
[3] Xiao Boxiang,Zhang Qiang,Wei Xiaopeng.Survey on Human Mocap Data Feature Extraction and Retrieval.Application Research of Computers,2010,27(1): 10-13(in Chinese)
(肖伯祥,张强,魏小鹏.人体运动捕捉数据特征提取与检索研究综述.计算机应用研究,2010,27(1): 10-13)
[4] Wang Manjun.Research on Virtual Human Motion Synthesis Techniques and Engineering Application.Ph.D Dissertation.Changsha,China: National University of Defense Technology,2010: 5-10(in Chinese)
(王满军.虚拟人运动合成技术及其工程应用研究.博士学位论文.长沙:国防科技大学,2010: 5-10)
[5] Liu Feng.Techniques for Motion Capture Data Based Animation.Ph.D Dissertation.Hangzhou,China: Zhejiang University,2004: 29-35(in Chinese)
(刘丰.基于运动捕获数据的若干动画技术研究.博士学位论文.杭州:浙江大学,2004: 29-35)
[6] Pan Hong,Xiao Jun,Wu Fei,et al.3D Human Motion Retrieval Based on Key-Frames.Journal of Computer-Aided Design Computer Graphics,2009,21(2): 214-222(in Chinese)
(潘红,肖俊,吴飞,等.基于关键帧的三维人体运动检索.计算机辅助设计与图形学学报,2009,21(2): 214-222)
[7] Xiao Jun.Intelligent Techniques for Character Animation.Ph.D Dissertation.Hangzhou,China: Zhejiang University,2007: 68-71(in Chinese)
(肖俊.智能人体动画若干关键技术研究.博士学位论文.杭州: 浙江大学,2007: 68-71)
[8] Zhang Shun,Liang Xiubo,Li Qilei.Performance-Driven Motion Retrieval and Its Usability Evaluation.Journal of Computer-Aided Design Computer Graphics,2009,121(4): 554-560 (in Chinese)
(张顺,梁秀波,李启雷,等.示范表演驱动的运动数据检索方法及可用性评估.计算机辅助设计与图形学学报,2009,121(4): 554-560)
[9] Wang Xing,Yu Zhiwen,Wang H S.Searching of Motion Database Based on Hierarchical SOM // Proc of the IEEE International Conference on Multimedia and Expo.Hannover,Germany,2008: 1233-1236
[10] Okajima S,Okada Y.Hierarchical Visual Motion Retrieval System and Its Motion Features // Proc of the International Conference on P2P,Parallel,Grid,Cloud and Internet Computing.Fukuoka,Japan,2010: 90-97
[11] Meinard M,Tido R,Michael C.Efficient Content-Based Retrieval of Motion Capture Data.ACM Trans on Graphics,2005,24(3): 677-685
[12] Demuth B,Roder T,Muller M,et al.An Information Retrieval System for Motion Capture Data // Proc of the 28th European Conference on Information Retrieval.London,UK,2006: 373-384
[13] Meinard M,Tido R.Motion Templates for Automatic Classification and Retrieval of Motion Capture Data // Proc of the Eurographics/ACM SIGGRAPH Symposium on Computer Animation.Vienna,Austria,2006: 137-146
[14] Keogh E J,Pazzani M J.Scaling up Dynamic Time Warping to Massive Dataset // Proc of the 3rd European Conference on Principles of Data Mining and Knowledge Discovery.Prague,Czech Republic,1999: 1-11
[15] Keogh E,Palpanas T,Zordan V B,et al.Indexing Large Human-Motion Databases // Proc of the 30th International Conference on Very Large Data Bases.Toronto,Canada,2004: 780-791
[16] Lucas K.Automated Methods for Data-Driven Synthesis of Realistic and Controllable Human Motion.Ph.D Dissertation.Madison,USA: University of Wisconsin at Madison,2004: 1-163
[17] Lucas K,Michael G,Frederic P.Motion Graph.ACM Trans on Graphics,2002,21(3): 473-482
[18] Wang Zhihua.Time Series Matching: A Multi-Filter Approach.Ph.D Dissertation.New York,USA: New York University,2006: 10-12