基于骨骼的三维虚拟人运动合成方法研究

J4 ›› 2014, Vol. 36 ›› Issue (04): 737-740.

基于骨骼的三维虚拟人运动合成方法研究

何长鹏，侯进，王献

(西南交通大学信息科学与技术学院，四川成都 610031)

收稿日期:2012-09-04 修回日期:2012-12-19 出版日期:2014-04-25 发布日期:2014-04-25
基金资助:
国家自然科学基金资助项目（60902023，61171096）；2009年高等学校博士学科点专项科研基金资助项目(20090184120020，20090184120022)；中央高校基本科研业务费专项资金资助项目（SWJTU12CX093）；四川省动漫研究中心2012年度科研项目（DM201204）

Research of the threedimensional virtual human motion
synthesis method based on skeleton

HE Changpeng,HOU Jin,WANG Xian

(School of Information Science and Technology,Southwest Jiaotong University,Chengdu 610031,China)

Received:2012-09-04 Revised:2012-12-19 Online:2014-04-25 Published:2014-04-25

摘要/Abstract

摘要：

针对虚拟人运动合成中建立的人体模型存在复杂化、合成的虚拟人运动序列逼真度差的问题，提出了一种基于骨骼的虚拟人运动合成方法。在分析人体结构的基础上，通过三维图形软件获取人体骨骼数据，构建虚拟人体的骨骼模型。另外，将关键帧四元数球面插值算法与时间和空间变形方法相结合，生成多样化的虚拟人运动序列。实验结果验证了该方法的有效性。

关键词: 虚拟人, 运动合成, 关键帧插值, 时间空间变形

Abstract:

In view of the problem that the construction of virtual human model is complicated and synthetic virtual human movement sequences have poor fidelity, proposes a method of the virtual human motion synthesis based on skeleton. Based on the analysis of human body structure, the human bone data are acquired through threedimensional graphics software. In addition, quaternions spherical interpolation method and the time and space deformation of motion fusion method are combined to generates diversity virtual human movement sequences. The effectiveness of this method is verified through the experiment.

Key words: virtual human;motion synthesis;keyframe interpolation;time and space warping

何长鹏，侯进，王献. 基于骨骼的三维虚拟人运动合成方法研究[J]. J4, 2014, 36(04): 737-740.

HE Changpeng,HOU Jin,WANG Xian. Research of the threedimensional virtual human motion
synthesis method based on skeleton [J]. J4, 2014, 36(04): 737-740.

参考文献

［1］Hou Jin,Wu Ling,Xu Fang,et al.Realtime text driven avatars［J］. Online Journal on Computer Science and Information Technology, 2011, 1(1):914.
［2］Peinado M,Meziat D,Maupu D,et al.Fullbody avatar control with environment awareness［J］. IEEE Computer Graphics and Applications, 2009, 29(3):6275.
［3］Zhang Guijuan, Zhu Dengming, Qiu Xianjie, et al. Skeletonbased control of fluid animation［J］. Visual Computer, 2011, 27(3):199210.
［4］Li Shilei, Liang Jiahong, Li Meng, et al. Whole body posture generation of virtual characters by combinig posture examples and prioritized inverse kinematics［J］. Journal of System Simulation, 2011, 23(12):26922700.(in Chinese)
［5］Kong Dehui, Wang Lichun, Zheng Chongyu. A key frame interpolation method enhancing motion details of skeletal animation［J］. Journal of Beijing University of Technology, 2011, 37(8):12551260.(in Chinese)
［6］Meng Wei, Guo Weibin, Zhu Changming, et al. Motion fusion technique based on characteristics of human motion［J］. Journal of Computer Applications, 2011, 31(1):4244.(in Chinese)
［7］Horvath C,Geiger W.Directable, highresolution simulation of fire on the GPU［C］∥Proc of the 2009 Computing Machinery’s Special Interest Group on Computer Graphics, 2009:18.
［8］Yuan Peng, Wang Shujie, Zhang Jiwei, et al. Virtual reality platform based on open sourced graphics toolkit openscenegraph［C］// Proc of the 10th IEEE International Conference on ComputerAided Design and Computer Graphics, 2007:361364.
附中文参考文献：
［4］李石磊, 梁加红, 李猛, 等. 基于姿态库和PIK算法的虚拟人全身姿态生成［J］. 系统仿真学报, 2011, 23(12):26922700.
［5］孔德慧, 王立春, 郑重雨. 增强骨骼动画运动细节的关键帧插值方法［J］. 北京工业大学学报, 2011, 37(8):12551260.
［6］孟巍, 郭卫斌, 朱昌明, 等. 基于人体运动规律的运动融合技术［J］. 计算机应用, 2011, 31(1):4244.

[1]	陈静杰，耿丽琴. 基于启发式Bi-RRT算法的虚拟手臂运动规划[J]. 计算机工程与科学, 2019, 41(03): 531-537.
[2]	戚福洲1，侯进1, 2，黄永坤1. 一种基于国标人体测量学数据的虚拟人建模方法[J]. J4, 2015, 37(04): 783-789.
[3]	胡维华, 钱克宠. 三维跑步场景与多角色虚拟人的融合设计[J]. J4, 2010, 32(3): 52-54.
[4]	方琦峰康凤举王宝龙王娇艳. 基于DI-Guy的人体视景仿真研究与应用[J]. J4, 2008, 30(1): 60-62.
[5]	薛志东[1,3] 隋卫平[2] 王乘[3]. 一种分割虚拟人彩色冰冻切片图像的混和方法[J]. J4, 2006, 28(9): 60-63.
[6]	卢晓军李焱贺汉根. 一种基于自由空间法的虚拟人行走规划方法[J]. J4, 2005, 27(8): 60-61.
[7]	卢晓军李焱贺汉根. 一种用于动态环境的虚拟人行走规划方法[J]. J4, 2004, 26(10): 75-78.