基于点云数据的人体特征尺寸提取
|
摘要
随着三维非接触人体扫描技术的出现,可以在几秒到几十秒内得到整个人体信息的三维扫描数据。这项技术已经在许多领域里被应用,但它仍是一项新技术,在服装工业领域中有很大的应用空间,服装业的网络量身定制(made to measure以下简称MTM)、三维虚拟设计、虚拟试衣等都依赖于三维扫描技术的发展。目前已开发使用的三维扫描设备可以提供高质量的人体模型,但多数测量设备提供的仅为没有特征信息的点云数据,不能直接得到3D服装CAD所需要的特征尺寸信息,而三维人体扫描得到的人体三维点一般为300,000多个,其中隐含了大量的信息,所以三维人体测量技术和尺寸信息提取技术是整个三维服装CAD技术发展的基础。为了使三维扫描数据能够被后续过程应用,必须对人体的各个特征点及特征尺寸进行提取。
本文主要论述通过三维扫描仪获取真实人体点云数据后,对非逻辑的点云数据进行数据识别、分割,结合人体的几何特征以及人体特征尺寸在服装学中的定义获取特征尺寸的测量基准点或特征点所在位置,过这些测量基准点或特征点获取截面,通过B样条曲线拟合截面上的数据点获取特征尺寸;阐述了扫描人体正背面的识别、肢体的识别和划分、关键测量基准点和特征点的识别、特征尺寸提取等方法,得出了服装应用中的主要人体尺寸如颈围、胸围、腰围、臀围、肩宽、背长、臂根围等特征尺寸。
With the appearance of the 3d non-contact body scanning technology, the whole human 3D data can be gotten in a time from seconds to dozens of seconds. 3D non-contact body scanning technology has been used in many fields, but it is still a new technology in the apparel industry, and there has a big space for it to apply. The development of made to measure in the apparel industry,3D virtual design, and virtual fitting all depend on the development of 3D scanning technology. High quality model of the human body can be provided by the 3D scanning equipment which has already been developed at present. But most of the measuring equipments got just the point cloud data without the characteristic information. The information of feature measurement which 3D garment CAD need can’t be gotten directly. However 3D body scanning can get the 3D points more than 300,000, which involves a large amount of information. So the technologies of 3D body measurement and extracting human body feature measurement are the base of the whole 3D garment CAD technology development. In order to make the 3D scanning data be put into use in the follow-up process, the feature points and feature measurement must be extracted.
This paper mainly discussed that identification and division of the logical point cloud data after point cloud data of body were obtained through the 3D scanner. Combing with the definitions of the geometrical features and feature parts of 3D body in Clothology, got the datum points or feature points of feature measurement. According to these datum points or feature points obtained the data of section which was corresponding to the feature measurement. B-spline was taken to fit these data in order to get the feature measurement. This paper expounded the methods of the recognition of the front and back of scanned human body, the identification of the scanned human body, and recognition of the datum points and feature point, extracted feature measurement and so on. Finally, the feature measurement of neck circumference, chest circumference, waistline, hipline, shoulder, neck to waist and armhole, etc , in the clothing application were gotten.
本文主要论述通过三维扫描仪获取真实人体点云数据后,对非逻辑的点云数据进行数据识别、分割,结合人体的几何特征以及人体特征尺寸在服装学中的定义获取特征尺寸的测量基准点或特征点所在位置,过这些测量基准点或特征点获取截面,通过B样条曲线拟合截面上的数据点获取特征尺寸;阐述了扫描人体正背面的识别、肢体的识别和划分、关键测量基准点和特征点的识别、特征尺寸提取等方法,得出了服装应用中的主要人体尺寸如颈围、胸围、腰围、臀围、肩宽、背长、臂根围等特征尺寸。
With the appearance of the 3d non-contact body scanning technology, the whole human 3D data can be gotten in a time from seconds to dozens of seconds. 3D non-contact body scanning technology has been used in many fields, but it is still a new technology in the apparel industry, and there has a big space for it to apply. The development of made to measure in the apparel industry,3D virtual design, and virtual fitting all depend on the development of 3D scanning technology. High quality model of the human body can be provided by the 3D scanning equipment which has already been developed at present. But most of the measuring equipments got just the point cloud data without the characteristic information. The information of feature measurement which 3D garment CAD need can’t be gotten directly. However 3D body scanning can get the 3D points more than 300,000, which involves a large amount of information. So the technologies of 3D body measurement and extracting human body feature measurement are the base of the whole 3D garment CAD technology development. In order to make the 3D scanning data be put into use in the follow-up process, the feature points and feature measurement must be extracted.
This paper mainly discussed that identification and division of the logical point cloud data after point cloud data of body were obtained through the 3D scanner. Combing with the definitions of the geometrical features and feature parts of 3D body in Clothology, got the datum points or feature points of feature measurement. According to these datum points or feature points obtained the data of section which was corresponding to the feature measurement. B-spline was taken to fit these data in order to get the feature measurement. This paper expounded the methods of the recognition of the front and back of scanned human body, the identification of the scanned human body, and recognition of the datum points and feature point, extracted feature measurement and so on. Finally, the feature measurement of neck circumference, chest circumference, waistline, hipline, shoulder, neck to waist and armhole, etc , in the clothing application were gotten.
引文
[1] Danen,Brunsman,Robinette. Reduing movement artifacts in whole bodyscanning[J]. 1997, IEEE :262.
[2]李燕.基于Pro/E的人体特征尺寸提取[J].工程图学学报,2009(1):12-16.
[3]徐军,何梅娟.三维服装CAD中关键技术的特征和方法[J].天津工业大学学报,2007:26(5):52-57.
[4]周旭东,李艳梅.人体三维测量技术分析[J].上海纺织科技,2002(6):58-59.
[5]肖正扬,陶恒学,杨继新等.基于数字化立体裁剪的人体非接触式测量技术[J].组合机床与自动化加工技术,2001(11):45-47.
[6]王祺明.服装业三维人体测量技术的方法和现状分析[J].绍兴文理学院学报,2002(4):66-69.
[7] SIMMONS K P.三维身体扫描仪应用潜力的比较[J].王祺明译.国外纺织技术,2002(7):35-37.
[8]刘卉,许瑞清,陈纯.服装CAD综述[J].计算机辅助设计与图形学学报,2000(6):476-478.
[9] Meadows D M,JOHNSON W O,WLLEN J B. Generation of surface contours by more patterns [J]. Applied Optics, 1970,9(4):942-948.
[10]李勇,付小莉,尚会超.三维人体测量方法的研究[J].纺织学报,2001(4):261-263.
[11]胡志萍,肖正扬,陶学恒等.服装立体设计中的非接触测量技术[J].大连轻工业学院学报,2001(1):44-48.
[12]王琦明.服装业三维人体测量技术的方法和现在分析[J].绍兴文理学院学报,2002(4):65-68.
[13]甘应进,陈东升,孟爽等.非接触式三维人体计测现状[J].纺织学报, 2005, 26(3):145-146.
[14] G. R. Geisen, C. P. Mason, V. L. Houston, J. J. Whitestone, B. K. Mc-Quiston, and A.C.Beattie. Automatic detection,identification,and registration of anatomical landmarks[J]. Proc.Human Factors Ergonom.Soc,v01.2:750—753,1995.
[15] E. A. Lewark and J. H. Nurre. Automated fidueials labeling on humanbody data[J] . Proc.1S&T/SPIE Symp.Electron.Imaging Sci.Technol.,vol.3313,San Jose,CA,Jan.1998.
[16] L. Dekker and P. Treleaven. Learning sysems for automated anthropometric landmarking[J]. Proc.Workshop 3D Anthropometry Industr.Products Design, Paris, France, 1998:111-116.
[17] Charlie C. L. Wang, Terry K.K.Chang, Matthew M.F.Yuen. From laser-scanned data to feature human model a system based on fuzzy logic concept[J]. Computer-Aided Design vol.35 2003: 241-253.
[18] M.H.Demers, J.D.Hurley, R.C.Wulpern, and J. R. Grindon. Three dimensional surface capture for body measurement using projected sinusoidal patterns[J]. in Proc. IS&T/SPIE Symp. Electron.Imaging:Sei.Technol.,vol,3023,San Jose,CA,1997:13-25.
[19]吴众山.基于B样条的三维人脸曲面生成及特征提取研究[D].厦门:厦门大学2008.
[20]林德静,孙晓东.基于三维扫描的人体尺寸提取技术[J]. 2005(3):36-41.
[21]张颖,邹奉元.三维人体测量技术的原理及应用[J].浙江工程学院学报,2002,20(4):311-313.
[22]尧燕玲,陈永强,陈君.基于人体切片的人体特征尺寸提取[J].山东纺织经济,2010,5:80-82.
[23]黄吉兰.基于特征点的人体建模过程与实现[D].西南交通大学,2010.
[24]丁玉兰.人体工程学[M].北京理工大学出版社,1992.
[25] Y. Huang, W. Yu, T. Chen. Three-dimensional body scanning system for apparel mass customization[J]. Optical engineering,2002,41(7) :1475-1479.
[26]李晓东.二维及多维链表及其算法实现[J] .佛山科学技术学院学报,2003,21(3):35-38.
[27]严蔚敏,吴伟.数据结构(C语言版)[M].北京:清华大学出版社,1997:30-32.
[28]杨跃武,李晓东,雷晓平等.多维链表的递归实现算法研究[J].计算机工程与应用,2005 41(9):68-69.
[29]李雪飞,周毅灵.个性化三维人体OBJ模型文件的构建[J].纺织学报,2008,10(29):127-131.
[30]孙晓东.三维扫描人体的尺寸自动提取技术[D].北京:北京服装学院2006.
[31] KARLA P S, ISTOOK C . A Comparison of three-dimensional body scanning and physicalanthropometric methods [J]. Seoul KSCT/ITAA Joint World Conference held in Seoul. 2001,(3): 18-24.
[32]余国兴.女装结构设计与应用[M].上海:中国纺织大学出版社,1998.
[33] Dr, Roy P Pargas. Automating information extraction from 3D scan data[D] . Clemson University.
[34]林德静,孙晓东.基于三维扫描的人体尺寸提取技术[J].北京服装学院报2005(9):36-41.
[35]杨倩.基于点云数据的人体建构[D].北京:北京服装学院2009.
[36]闫迷军,宁涛,张兆璞.服装CAD中三维人体尺寸信息提取技术的研究[J] .工程图学学报,2005(1) :26-29.
[37]刘瑞璞.服装纸样设计原理与应用[M].中国纺织出版社,2008.
[38]刘雁,耿兆丰等.三维人体及服装建模方法分析与实现[J].计算机应用与软件,2004,21(10) :47-48.
[39]王家文等. MATLAB7.0编程基础[M].北京:机械工业出版社,2005.
[40]徐东艳,孟晓刚. MATLAB函数库查询辞典[M].北京:中国铁道出版社, 2006.
[41] Schoenberg I J . Contributions to the Problem of Approximation of Equidistant Data by Analytic Functions [ J ] . Quart Appl Maths ,1946 , (4) :45-99 ,112-141. [42 ] Curry H B ,Schoenberg I J . On Spline Distributions and Their Limits :The Polya Distribution Functions [J ] . American Mathematical Soc ,1947 ,53 :1114.
[43] De Boor C. On Calculating with B2Spline[J]. J Approx Theory ,1972 , (6) :50-62. [44 ] De Boor C. A Practical Guide of Splines [M]. Newyork Springer,1978. [45 ] Cox M. G. The numerical Evaluation of B-Splines[J ]. JINST Maths Application ,1972 ,10 :134-149.
[46]王增波,彭仁忠,宫兆刚. B样条曲线生成原理及实现[J] .石河子大学学报(自然科学报),2009 2(1):118-121.
[47]侯魏.基于胸部形态分类的三维钢圈设计[D].北京:北京服装学院,2010.
[48]邬红芳.人体尺寸测量CAT及服装CAD系统[ J] .安徽机电学院学报, 2000, 15( 4) : 29-31.
[2]李燕.基于Pro/E的人体特征尺寸提取[J].工程图学学报,2009(1):12-16.
[3]徐军,何梅娟.三维服装CAD中关键技术的特征和方法[J].天津工业大学学报,2007:26(5):52-57.
[4]周旭东,李艳梅.人体三维测量技术分析[J].上海纺织科技,2002(6):58-59.
[5]肖正扬,陶恒学,杨继新等.基于数字化立体裁剪的人体非接触式测量技术[J].组合机床与自动化加工技术,2001(11):45-47.
[6]王祺明.服装业三维人体测量技术的方法和现状分析[J].绍兴文理学院学报,2002(4):66-69.
[7] SIMMONS K P.三维身体扫描仪应用潜力的比较[J].王祺明译.国外纺织技术,2002(7):35-37.
[8]刘卉,许瑞清,陈纯.服装CAD综述[J].计算机辅助设计与图形学学报,2000(6):476-478.
[9] Meadows D M,JOHNSON W O,WLLEN J B. Generation of surface contours by more patterns [J]. Applied Optics, 1970,9(4):942-948.
[10]李勇,付小莉,尚会超.三维人体测量方法的研究[J].纺织学报,2001(4):261-263.
[11]胡志萍,肖正扬,陶学恒等.服装立体设计中的非接触测量技术[J].大连轻工业学院学报,2001(1):44-48.
[12]王琦明.服装业三维人体测量技术的方法和现在分析[J].绍兴文理学院学报,2002(4):65-68.
[13]甘应进,陈东升,孟爽等.非接触式三维人体计测现状[J].纺织学报, 2005, 26(3):145-146.
[14] G. R. Geisen, C. P. Mason, V. L. Houston, J. J. Whitestone, B. K. Mc-Quiston, and A.C.Beattie. Automatic detection,identification,and registration of anatomical landmarks[J]. Proc.Human Factors Ergonom.Soc,v01.2:750—753,1995.
[15] E. A. Lewark and J. H. Nurre. Automated fidueials labeling on humanbody data[J] . Proc.1S&T/SPIE Symp.Electron.Imaging Sci.Technol.,vol.3313,San Jose,CA,Jan.1998.
[16] L. Dekker and P. Treleaven. Learning sysems for automated anthropometric landmarking[J]. Proc.Workshop 3D Anthropometry Industr.Products Design, Paris, France, 1998:111-116.
[17] Charlie C. L. Wang, Terry K.K.Chang, Matthew M.F.Yuen. From laser-scanned data to feature human model a system based on fuzzy logic concept[J]. Computer-Aided Design vol.35 2003: 241-253.
[18] M.H.Demers, J.D.Hurley, R.C.Wulpern, and J. R. Grindon. Three dimensional surface capture for body measurement using projected sinusoidal patterns[J]. in Proc. IS&T/SPIE Symp. Electron.Imaging:Sei.Technol.,vol,3023,San Jose,CA,1997:13-25.
[19]吴众山.基于B样条的三维人脸曲面生成及特征提取研究[D].厦门:厦门大学2008.
[20]林德静,孙晓东.基于三维扫描的人体尺寸提取技术[J]. 2005(3):36-41.
[21]张颖,邹奉元.三维人体测量技术的原理及应用[J].浙江工程学院学报,2002,20(4):311-313.
[22]尧燕玲,陈永强,陈君.基于人体切片的人体特征尺寸提取[J].山东纺织经济,2010,5:80-82.
[23]黄吉兰.基于特征点的人体建模过程与实现[D].西南交通大学,2010.
[24]丁玉兰.人体工程学[M].北京理工大学出版社,1992.
[25] Y. Huang, W. Yu, T. Chen. Three-dimensional body scanning system for apparel mass customization[J]. Optical engineering,2002,41(7) :1475-1479.
[26]李晓东.二维及多维链表及其算法实现[J] .佛山科学技术学院学报,2003,21(3):35-38.
[27]严蔚敏,吴伟.数据结构(C语言版)[M].北京:清华大学出版社,1997:30-32.
[28]杨跃武,李晓东,雷晓平等.多维链表的递归实现算法研究[J].计算机工程与应用,2005 41(9):68-69.
[29]李雪飞,周毅灵.个性化三维人体OBJ模型文件的构建[J].纺织学报,2008,10(29):127-131.
[30]孙晓东.三维扫描人体的尺寸自动提取技术[D].北京:北京服装学院2006.
[31] KARLA P S, ISTOOK C . A Comparison of three-dimensional body scanning and physicalanthropometric methods [J]. Seoul KSCT/ITAA Joint World Conference held in Seoul. 2001,(3): 18-24.
[32]余国兴.女装结构设计与应用[M].上海:中国纺织大学出版社,1998.
[33] Dr, Roy P Pargas. Automating information extraction from 3D scan data[D] . Clemson University.
[34]林德静,孙晓东.基于三维扫描的人体尺寸提取技术[J].北京服装学院报2005(9):36-41.
[35]杨倩.基于点云数据的人体建构[D].北京:北京服装学院2009.
[36]闫迷军,宁涛,张兆璞.服装CAD中三维人体尺寸信息提取技术的研究[J] .工程图学学报,2005(1) :26-29.
[37]刘瑞璞.服装纸样设计原理与应用[M].中国纺织出版社,2008.
[38]刘雁,耿兆丰等.三维人体及服装建模方法分析与实现[J].计算机应用与软件,2004,21(10) :47-48.
[39]王家文等. MATLAB7.0编程基础[M].北京:机械工业出版社,2005.
[40]徐东艳,孟晓刚. MATLAB函数库查询辞典[M].北京:中国铁道出版社, 2006.
[41] Schoenberg I J . Contributions to the Problem of Approximation of Equidistant Data by Analytic Functions [ J ] . Quart Appl Maths ,1946 , (4) :45-99 ,112-141. [42 ] Curry H B ,Schoenberg I J . On Spline Distributions and Their Limits :The Polya Distribution Functions [J ] . American Mathematical Soc ,1947 ,53 :1114.
[43] De Boor C. On Calculating with B2Spline[J]. J Approx Theory ,1972 , (6) :50-62. [44 ] De Boor C. A Practical Guide of Splines [M]. Newyork Springer,1978. [45 ] Cox M. G. The numerical Evaluation of B-Splines[J ]. JINST Maths Application ,1972 ,10 :134-149.
[46]王增波,彭仁忠,宫兆刚. B样条曲线生成原理及实现[J] .石河子大学学报(自然科学报),2009 2(1):118-121.
[47]侯魏.基于胸部形态分类的三维钢圈设计[D].北京:北京服装学院,2010.
[48]邬红芳.人体尺寸测量CAT及服装CAD系统[ J] .安徽机电学院学报, 2000, 15( 4) : 29-31.