纺织学报 ›› 2012, Vol. 33 ›› Issue (8): 149-154.

• 管理与信息化 • 上一篇    下一篇

基于多分辨率椭球包围盒的动态服装仿真

高军1,李重1,2,林佼1,李亮1   

    1. 浙江理工大学信息电子学院
    2. 浙江理工大学服装实验教学中心
  • 收稿日期:2011-10-11 修回日期:2012-02-11 出版日期:2012-08-15 发布日期:2012-08-08
  • 通讯作者: 李重 E-mail:lizhongzju@hotmail.com
  • 基金资助:

    国家自然科学基金资助项目;浙江省自然科学基金资助项目;浙江省钱江人才D类项目;国家级实验教学示范中心项目;全国高等学校教学研究中心项目;浙江理工大学研究生教改项目

Dynamic garment simulation based on multi-resolution ellipsoid bounding box

 GAO  Jun1, LI  Zhong1,2, LIN  Jiao1, LI  Liang1   

    1. College of Information & Electronics, Zhejiang Sci-Tech University
    2. Garment Experimental Teaching Center, Zhejiang Sci-Tech University
  • Received:2011-10-11 Revised:2012-02-11 Online:2012-08-15 Published:2012-08-08
  • Contact: LI Zhong E-mail:lizhongzju@hotmail.com

摘要: 服装动态仿真中,服装与人体模型之间的碰撞检测与响应是影响仿真速度的一个重要因素。为了提高服装仿真的效率和效果,本文提出了一种基于多分辨率椭球包围盒的快速服装仿真方法。在人体模型八叉树体素分解基础上,进行了球体整合和优化。为了降低被检测球体的数量,对球形包围盒进行了椭球体融合,形成了一系列多分辨率的模型近似。并通过交互式方法设计服装衣片的缝合,使得服装模拟具有更大自由度。实验结果表明,该方法仿真真实感强,对改进服装动态仿真的效率有很大帮助。

关键词: 球形包围盒 , 体素划分 , 八叉树 , 多分辨率 , 椭球体

Abstract: The collision detection and response between the garment and the human body model is an important factor when dealing with dynamic clothing simulation. In order to improve the efficiency and effect of the simulation, this paper presents a new method based on the ellipsoid bounding box in multi-resolution. After subdividing the human model by using octree, we first reorganize and optimize the voxels. Then we merge the sphere bounding boxes based on the multi-resolution ellipsoids for accelerating the collision detection, and form a series of model approximations. We also interactively design the garment stitching, which makes the cloth simulation more flexible. The experimental result shows that the simulation is more realistic and it is helpful to improve the efficiency of the cloth simulation.

Key words: sphere bounding box , voxel division , octree , muoti-resolution , ellipsoid

中图分类号: 

  • TP 391
[1] 景军锋 李阳 李鹏飞 焦洋. 基于小波域多尺度Markov随机场的织物印花图案分割[J]. 纺织学报, 2014, 35(1): 127-0.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!