三维人体信息获取及虚拟服装试穿技术研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
现代数字化服装技术的发展使服装企业现在能够以较低的成本收集、分析不同客户
    的资料和需求,通过灵活、柔性的生产系统分别定制。大规模量体定制生产方式将给每
    个客户带来个性化的商品和服务成为可能。构建一个基于大规模量体定制技术的电子商
    务平台,既可以面向顾客提供虚拟购物,个性化的合身服装等高附加值服务,同时也可
    提高企业的制造效率,降低企业的运行成本,增强企业的竞争力。
    考虑国内服装行业的现状与国际服装行业发展的趋势,结合已有的科研基础,进行
    了三维全身扫描仪的研制,并对与之相关,且同为大规模量身定制关键技术的人体尺寸
    自动获取与三维人体表面模型构建,以及虚拟服装试穿技术进行了研究。具体内容如下:
    (1)概要性的介绍了基于大规模量体定制技术的电子商务平台的基本框架,并论述
    三维全身扫描仪,基于三维人体扫描数据的三维人体测量技术及服装穿着效果仿真技术
    的国内外研究概况。
    (2)针对三维人体扫描中的技术问题,采用了合理的设计方案与算法实现三维全身
    扫描。三维全身扫描仪使用线激光源,并采用双摄像机互补,多扫描头数据融合的方法
    克服人体自遮挡,获取人体表面的完整点云数据。激光条纹中心提取的准确性直接影响
    到系统的测量精度,针对人体肤色与环境光照的变化,采用基于遗传算法的自适应阈值
    法提取激光条纹中心线,该方法具有适应性强,更加准确的优点。摄像机的定标采用基
    于结构光平面约束的 DLT 模型,并对双相机图像的映射关系进行了推导,给出了双相机
    定标的算法。依据三维全身扫描方案获得的点云数据的拓扑结构有其特殊性。首先,对
    扫描线进行最小二乘 B 样条曲线拟合处理,消除采样误差,提高测量精度。然后,对拟
    合的 B 样条曲线重新等弧长采样。最后,采用一种高效、准确的方法——“加权距离法”,
    融合相对应的扫描线。三维人体扫描仪的扫描结果验证了该三维全身扫描仪系统结构与
    三维信息获取方案的合理性。
    (3)提出了一种基于 Open Inventor 软件开发平台的结构光测距系统仿真方法。从
    仿真系统中获得的扫描图像序列用于三维重建,可精确获取扫描对象的三维表面信息,
    实验结果验证了该仿真方法的有效性与实用性。并将该仿真方法用于三维全身扫描仪的
    
    
    系统设计。
    (4)提出了一个在分割人体扫描点云的基础上自动提取人体测量尺寸的方案。首先,
    通过计算人体扫描点云部分截面的一阶矩与二阶矩确定人体的朝向。其次,采用计算几
    何与图像处理的算法分割人体点云。第三,估计人体特征标记点的高度,通过查找局部
    特征确定标记点。最后,根据标记点计算人体尺寸。该测量方案在测量的准确性与精确
    性上均符合使用要求。
    (5)人体的形状比较复杂,存在多个分支,从扫描得到的序列轮廓线重构人体表面
    会存在空洞。可将多轮廓线间的形体重构问题转化为体数据中的等值面构造问题,通过
    定义场函数将一系列平面上的多条轮廓线就转换为体数据,并采用 Marching Cubes 方
    法抽取等值面,最终获得三维人体表面模型,完成三维人体表面重建。采用该方法重建
    的人体表面完整,光滑且符合精度要求。
    (6)引入弹簧-质点模型,对质点的受力情况进行了详尽的分析,并给出了基于显
    式的欧拉方法的变形模型的求解,在此基础上实现了布料仿真。采用了基于正则栅格法
    的四边域网格剖分算法以实现对二维衣片的离散,剖分生成的网格结构充分满足二维衣
    片转三维服装的特殊要求。在给定衣片之间缝合对应关系的前提下,给出两种最主要的
    缝合模式:对位缝合和抽褶缝合。碰撞检测是整个模拟过程的关键。通过构建人体模型
    的方向包围盒树可以大大提高碰撞检测的效率。并采用直接修正质点运动状态(位置及
    速度)的方法解决响应问题。最后在三维虚拟服装试穿仿真程序中,实现了上述大部分
    算法。实验结果证明,这些方法是切实可行的,不仅很好地完成了平面衣片到立体服装
    的转换,而且真实自然地模拟出服装以及柔性织物的自然垂感。
The development of modern digital garment technology has made the enterprises be able
    to collect and analyze the information and requests of various customs at a low cost. Then the
    Mass Customization can be realized by the agile and flexible production system. The Mass
    Customization manner will make it possible to offer each custom the individual commodity
    and service. When an e-business platform based on mass customization technology is
    established, the custom-oriented virtual shopping and individual fitting garment are available.
    At the same time, the manufacture efficiency will be improved, together with the cost
    decrease, so as to enhance the competition of the enterprise.
     Considering the actuality of domestic garment industry and the trend of the international
    garment industry, we are developing the 3D whole body scanner on our basis. The correlative
    technologies, such as Automatic body measurement, Reconstructed 3D human body surface
    model, and Virtual Try-On, are studied in the paper. The details are as follow:
     (1) The e-business platform framework on the mass customization is introduced. Then
    we discuss the survey of 3D whole body scanner, 3D body measurement based on the scanned
    data sets and the Virtual Try-On simulation in and outside China.
     (2) To the problems in 3D body scanning, the rational design and algorithm are adopted
    to achieve the 3D whole body scanning. In the design, laser stripes are projected horizontally
    on the body. Then the methods of fusing the images of two cameras and integrating the data
    sets of scanning heads are adopted to overcome the body's occlusion so that we can get the
    complete contour lines of the body surface.
    The accuracy of structured light system hinges on the ability to locate the center of the laser
    stripe. To adapt the variation of skin and illumination, a method for detecting the center of
    laser stripe based on genetic algorithm (GA) is introduced for the purpose of improving the
    accuracy of the system. The mapping relationship between the images of two cameras is
    
    
    studied in this paper. Based on the planar-constraint DLT model, the calibration algorithm and
    mapping formula for the two cameras are presented. The topological characteristic of the data
    sets from our 3D whole body scanner is special. First, we fit the scanned lines with B-spline
    in order to eliminate the sampling error and improve the accuracy. Then the curves are
    resampled on equidistance. Finally the corresponding scanned lines are fused by an effective
    and exact method—weighed distance method. The scanned results validate the 3D whole
    body scanner design and the 3D information acquirement method.
     (3) The simulator of range finder based on structured light can be applied to optimal
    design and test. A simulator using the platform of Open Inventor has been proposed. The
    image sequences produced by the simulator are available in 3D reconstruction, and then the
    3D surface information of the subject is obtained accurately. The experimental result shows
    that the simulation method is effective and feasible. And apply this simulator in the system
    design of 3D whole body scanner.
     (4) A method of automatic body measurement based on segment human body data sets is
    proposed. First, Finding the location and direction of the subject is accomplished by
    calculating moments of the data. Second, Segmenting data by algorithms of computer
    geometry and image process. Third, estimating the high of the landmark and finding the
    landmark by local feature. At last, computing the measurement of human body by landmarks.
    The method accords with the requirement of human body measurement on the precision and
    the accuracy.
     (5) The sharp of human body is complex, which has a lot of branch. The reconstructed
    surface model of human body skin from the sequence of contours exists holes. The problem in
    construction of surface from contour data can be solved by the construction of Iso-surface in
    voxel data. The contour data in a series plans can be transformed voxel
引文
[1] Pine B J. Mass Customization: The New Frontier in Business Competition. Boston:
     Harvard Business School Press,1993
    [2] Pine B J, Victor B and Boynton A C. Making Mass Customization Work. Harvard
     Business Review, 1993,5(1): 108~119
    [3] Michael T Fralix, Cary N C. From Mass Production to Mass Customization. Journal of
     Textile and Apparel, Technology and Management, 2001,1(2):112~120
    [4] L McKinnon, C Istook. Consumer Acceptance of Body Scanning. Journal of Fashion
     Marketing and Management, 2001,3(2):78~89
    [5] C Istook. Automating Customization of Apparel. International Journal of e-Business
     Strategy Management,2002,9(5):34~41
    [6] C Istook. Technological Applications Enabling Mass Customization. International Journal
     of Clothing Science and Technology,2001,21(13):119~127
    [7] Frederic C, Hyewon S, Nadia M T.Made-to-Measure Technologies for an Online Clothing
     Store. IEEE Computer Graphics, 2003, 23(1): 38~48
    [8] James Clarke. E-TAILOR: Integration of 3D Body Measurement, Advanced CAD, and
     E-Commerce Technologies in the European Fashion Industry. International Journal of
     e-Business Strategy Management,2001,2(3):201~209
    [9] C Istook, K Simmons. Comparison Of Three-Dimensional Body Scanning Technologies.
     Journal of Fashion Marketing and Management, 2002,17 (13): 67~73
    [10] Paquette S. 3D scanning in apparel design and human engineering. IEEE Computer
     Graphics and Applications, 1996,16 (5):11~15
    [11] 丁玉兰,郭刚,赵江洪.人机工程学.北京:北京理工大学出版社,2000
    [12] E Paquet, K Robinette,M Rioux. Management of 3D and Anthropometric Databases:
     Alexandria and Cleopatra. Journal of Electronic Imaging, 2000,9(4): 91~95
    [13] 赖维铁.人机工程学.华中理工大学出版社. 1989
    [14] 赵欲晓.浅谈人体计测的发展.服装设计师,2002, 8(2):129~131
     122
    
    
    [15] Jones PRM, West G, Harris D. The Loughborough Anthropometric Shadow Scanner
     (LASS). Endeavour, 1989,13(4): 164~168
    [16] Hurley J D, Demers M H, Wulpern R C. Body measurement system using white light
     projected patterns for made-to-measure apparel. SPIE,1997,( 3131):212~223
    [17] K. Simmons, Body Measurement Techniques: A Comparison of Three-Dimensional
     Body Scanning and Physical Anthropometric Methods. ITAA/KSCT 2001 Conference,
     Seoul, KOREA,2001
    [18] Daanen H, Taylor SE, Brunsman M A. Absolute accuracy of the Cyberware WB4 whole
     body scanner. SPIE,1997,( 3023): 6~12
    [19] C E Clauser, I Tebbets, B Bradtmiller. Measurer’s Handbook: US. Army Anthropometric
     Survey. New York: Nation Press, 1988
    [20] Jones P R M, Rioux M. Three-dimensional Surface Anthropometry: Applications to the
     Human Body. Optics and Lasers in Engineering, 1997,28(9):89~117
    [21] R P Pargas, N J Staples, J S Davis. Automatic Measurement Extraction for Apparel from
     a Three-dimensional Body Scan. Opt. Engng, 1997,28(2):157~172
    [22] R P Pargas, N J Staples, B F Malloy et al. Tilted Planes in 3D Image Analysis. SPIE,
     1998,(3131):212~223
    [23] K Cantrell. Extracting Human Body Measurements from 3D Body Point Scans :[Masters
     paper]. Clemson University, 1997
    [24] D Burnsides. 3-D Landmark Detection and Identification in the CAESAR Project. Inter-
     national Conference of Recent Advances 3D Digital Imaging and Modeling, IEEE NJ
     USA, 2001. 393~398
    [25] Simmons K P and Istook C L. Body measurement techniques: A comparison of 3D body
     scanning and physical anthropometric methods for apparel application. Journal of Fashion
     Marketing and Management, 2003,7 (3): 306~332
    [26] D E Breen. Computer graphics in textiles and apparel modeling. IEEE Computer
     Graphics and Applications,1996,16(5):26~27
    [27] P. Volino, N. Thalmann. Developing simulation techniques for an interactive clothing
     system. IEEE International Conference on Virtual Systems and Multimedia, 1997. 109~
     123
    
    
    118
    [28] Tae J, Sung Min Kim. prediction of garment drape shape. International Journal of
     Clothing Science and Technology.2000,12(1):39~49
    [29] J Ascough. A simple finiteelement model for cloth drape simulation. International
     Journal of Clothing Science and Technology, 1996,8(3):59~74
    [30] Bathe K J.Finite Element Procedures in Engineering Analysis. Englewood Cliffs, NJ:
     Prentice-Hall, 1982
    [31] Breen DE, House D H, Wozny MJ. Predicting the drape of woven cloth using interacting
     particles. Proceedings of SIGGRAPH’94. Computer Graphics. 1994. 365~372
    [32] Weil J. The synthesis of cloth objects. Proceedings of SIGGRAPH’86. Computer
     Graphics, 1986,20(4): 49~54
    [33] Barzel R. Physically based Modeling for Computer Graphics. New York: Academic
     Press,1992
    [34] Matsuda R, Imaoka H. A graphic method to simulate garment fitting on a human body
     model in various posture. Seni Gakkaishi,1995,51(5):225~233
    [35] Provot X. Deformation constraints in a mass-spring model to describe cloth behavior. In:
     Wayne A ed. Proceedings of the Graphics Interface Conference'95. Vancouver. Canadian
     Human-Computer Communications Socity,1995.147~154
    [36] Okabe H, Imaoka H, Tomiha T et al.Three dimensional apparel CADsystem. Computer
     Graphics,1992,26(2):105~129
    [37] Vassilev T I . Dressing Virtual People.in: SCI'2000 conference. Orlando. 2000.124~132
    [38] M Ishii, T Nagata. Feature extraction of three-dimensional object and processing in
     hand-eye system using laser tracker. Pattern Recognition,1976,8(4):229~237
    [39] RA Jarvis. a Perspective on Range Finding Techniques for Computer Vision. IEEE PAMI,
     1983,5(2)122~139
    [40] O Ozeki. Real time range measurement device for 3D object recognition. IEEE PAMI,
     1986,8(5): 550~553
    [41] Brian Curless. From Range Scans to 3D Models. Computer Graphics,1999, 33(4):38~41
    [42] 周长发.精通 Visual C++.NET 图像处理编程.北京:电子工业出版社,2002
     124
    
    
    [43] 艾勇.最新光学应用测量技术.武汉:测绘科技大学出版社,1994
    [44] 吴剑波,崔振,赵宏等.光刀中心自适应阈值提取法.半导体光电,2001,22(1):62~64
    [45] 潘正君等.演化计算.北京:清华大学出版社,1998
    [46] 张文修,梁怡.遗传算法的数学基础.西安:西安交通大学出版社,2000
    [47] 章毓晋. 图像分割.北京:科学出版社,2001.103~107
    [48] 郑宏,潘励.基于遗传算法的图像阈值的自动选取.中国图像图形学报,1999,4
     (4):327~330.
    [49] 马颂德,张正友.计算机视觉-计算理论与算法基础. 北京:科学出版社,1998
    [50] R Klette, K Schlüns, A Koschan. Computer Vision: Three Dimensional Data from Images.
     Springer Press,1998
    [51] 金刚,叶建辉,李德华等.一种用于三维激光扫描仪的结构光测距方法和装置定标.红
     外与激光.1996,11(3): 61~63
    [52] 金刚,李德华等.双相机结构光系统中两相机定标与映射计算.华中科技大学学
     报.2002,30(10):62~64
    [53] 谷口庆治.数字图像处理(下). 朱虹,廖学成,乐静等译.北京:科学出版社,2002
    [54] 易大义,陈道琦.数值分析引论.杭州:浙江大学出版社,1998
    [55] G Turk, M Levoy. Zippered Polygon Meshes form Range Images. Proceedings of
     SIGGRAPH.1994,2(1):311~318
    [56] 吴重光.仿真技术.北京:化学工业出版社,2000
    [57] 顾启泰.系统设计与仿真.北京:清华大学出版社,1995
    [58] Open Inventor Architecture Group. Open Inventor ? C++ Reference Manual. Addison
     Wesley Publishing Company, 1994
    [59] 胡斌,李德华,金刚.基于方向模板的结构光条纹中心检测方法.计算机工程与应
     用.2002,32(6):59~60.
    [60] Eric Lengyel. The OpenGL Extensions Guide. Charles River Media, 2003
    [61] 金刚,吴晓刚,李德华.三维扫描系统中的关键技术.计算机世界,1999,18(7):11
    [62] 金刚,李德华,李泽宇.三维激光彩色扫描仪中物体表面色彩信息获取.华中理工大学
     学报,1999,27(7):95~97
     125
    
    
    [63] JIN Gang, Lidehua, Lizeyu. Acquiring depth information of object’s low reflectance
     surface in three dimensional laser color scanner. Proc. 9th National Conf. on Image and
     Graphics, pp241-243, Xi’an China, 1998.
    [64] Jiangyu Zheng. Acquiring 3-D models from sequences of contours. IEEE PAMI,
     1994,16(2):163~178
    [65] JIN Gang, YeJianhui, TangLijun et al. A range-finding technique and calibration
     algorithm use structured light for 3-D laser scanner. Proc. 8th National Conf. on Image
     and Graphics, pp660-663, Wuhan, China, 1996
    [66] Graham R L. An efficient algorithm for determining the convex hull of a finite planar set.
     Info Proc Lett, 1972,7(1):132~133
    [67] Gonzalez R C, Woods R E. Digital Image Processing, 2nd ed, Prentice Hall, 2002
    [68] 关治, 陈景良.数值计算方法.北京: 清华大学出版社, 1990
    [69] J P Marques de Sa. Pattern Recognition: Concepts, Methods and Applications. Springer
     Verlag Berlin and Heidelberg GmbH & CoK, 2001
    [70] Marc E Herniter. Programming in MATLAB. Brooks:Cole Publishing Company, 2001
    [71] American Standards for Testing and Materials (ASTM). Standard table of body
     measurements for women aged 55 and older (all figure type). Washington: West
     Conshohocken Press,1995
    [72] American Standards for Testing and Materials (ASTM). Standard terminology relating to
     body dimensions for apparel sizing. Washington: West Conshohocken Press, 1999
    [73] American Standards for Testing and Materials (ASTM). Standard table of body
     measurements for adult female misses figure type (Vol. 07-02, Designation: D5585-95).
     Washington: West Conshohocken Press, 1995
    [74] 帕夫利迪斯.结构模式识别.上海:上海科学技术文献出版社, 1981
    [75] 唐泽圣.三维数据场可视化.北京:清华大学出版社,1999
    [76] Jones M W, Chen M. A New Approach to the Construction of Surfaces from Contour
     Data. Eurographics’94. 1994,13(3):75~84
    [77] Lorensen W E, Cline H E. Marching Cubes: A High Resolution 3D Suface Construction
     Algorithm. Conputer Graphics. 1987,21(4):163~169
     126
    
    
    [78] Durst M J. Letters: Additional Reference to Marching Cubes. Computer Graphics.1988,
     18(2):22
    [79] Nielson G M, Foley T A, Hamann B, et al. Visualizing and Modeling Scattered
     Multivariate Data. IEEE Computer Graphics & Applications. 1991,9(5):47~55
    [80] Wilhelms J, Challinger J, Alper N, et al. Direct Volume Rendering of Curvilinear
     Volumes. Computer graphics. 1990,24(5):41~47
    [81] Chen M, Tang Z, Tang L. Generating Grouping Isosurfaces from Multi-resolution 3D
     Data sets. Proceedings of CAD&Graphics, 1997,47(33):262~267
    [82] Preparata F P, Shamos M I. Computational Geometry: An introduction. New York:
     Springer-Verlag, 1988
    [83] Francis S, Hill, Jr. Computer Graphics Using Open GL, Second Edition. The Prentice
     Hall Computer Science Team, 2000
    [84] David F Rogers, J Alan Adams. Procedural Elements of Computer Graphics.
     McGraw-Hill Science, 1989
    [85] Louchet J, Provot X. Evolutionary identification of cloth animation models. In:
     Terzopoulos D, Thalmann Deds. Proceedings of the Computer Animation and
     Simulation'95. New York: Springer-Verlag,1995.44~54
    [86] 樊劲,周济,王启付.基于弹簧质点模型的二维/三维映射算法.软件学报, 1999,
     10(2):140~148
    [87] Bathe K J. In Finite Element Procedures in Engineering Analysis. Prentice Hall, 1982
    [88] 聂卉.三维柔性织物动态仿真技术的研究:[博士学位论文].中山大学,2002
    [89] Chin S P, Wu C Y. A new grid-generation method for 2-D simulation of devices with
     nonplanar semiconductor surface. IEEE Computer-Aided Design of Integrated Circuits
     and Systems,1993,12(9):1337~1344
    [90] Chittaro L, Corvaglia D. 3D Virtual Clothing: from Garment Design to Web3D
     Visualization and Simulation, Proceedings of Web3D 2003: 8th International
     Conference on 3D Web Technology, ACM Press, New York, March 2003, 73~84
    [91] Michel Carignan, Ying Yang, Nadia Magnenat Thalmann et al. Dressing Ani- mated
     Synthetic Actors with Complex Deformable Clothes. Computer Graphics,1992, 26(2):
     127
    
    
    99~104
    [92] Provot X. Deformation Constraints in a Mass-Spring Model to describe Rigid Cloth
     Behavior. Institut National de Recherche en Informatique et Automatique (INRIA) Le
     Chesnay Cedex, France. 1995
    [93] Lin M C, Manocha D, Cohen J, et al. Collision Detection: Algorithms and Applications.
     University of North Carolina ,1996
    [94] 石教英.虚拟现实基础及实用算法.北京:科学出版社,2002
    [95] Moore M, Wilhelms J. Collision Detection and Response for Computer Anima- tion.
     Computer Graphics (SIGGRAPH’88 proceedings), Addison-Wesley,pp 289-298, 1988
    [96] Gottschalk S, Lin M C, Manocha D. OBBTree: A Hierarchical Structure for Rapid
     Interference Detection. Department of Computer Science,University of North
     Carolina,1996
    [97] Klosowski JT, Held M, Mitchell J S B. Efficient Collision Detection Using Bounding
     Volume Hierarchies of k-dops. IEEE Visualization and Computer Graphics,
     1997,10(7):1122~1338
    [98] X Provot. Collision and Self-collision Handling in Cloth Model Dedicated to Design
     Garments. Computer Graphics Interface. 1997, 12(6): 177~189
     128

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700