摘要
在医疗设备,服装设计,电影制作和电脑游戏中三维人体模型是不可或缺的元素,但是通过扫描等途径直接获取的人体模型在很多环境下并不能直接运用,因此需要对模型根据所处环境进行适度变形。研究三维人体模型变形方法,引入语义驱动模型的变形。通过测量模型骨骼关节点位置并得到身高、腿长及肩高等模型语义值,对模型进行PCA降维,建立模型语义值与模型矩阵之间的映射关系,调整相关语义值获得相应PCA权值,并且将PCA权值重新映射到模型矩阵得到变形后的模型。
In medical equipment, clothing design, film production and computer games, 3 D human body model is an indispensable element, but directly obtaining it by means of scanning the human body model does not directly apply in many environments, so we need to moderate deformation model according to the environment. The operation difficulty of the model of deformation is considered, and the deformation effect is achieved. For studying the 3 D human body model deformation problem, the deformation of semantic driven model is introduced. Semantic values of height, leg length and shoulder height were obtained by measuring the position of joint points in the model skeleton so as to adjust model related semantic values and get the deformation of the model.
引文
[1]Hasler N,Stoll C,Sunkel M,et al.A statistical model of human pose and body shape[C]//Computer Graphics Forum.Blackwell Publishing Ltd,2009,28(2):337-346.
[2]Gao L,Lai Y K,Huang Q X,et al.A Data‐Driven Approach to Realistic Shape Morphing[C]//Computer Graphics Forum.Blackwell Publishing Ltd,Spain 2013,2013,32(2pt4):449-457.
[3]Fr?hlich S,Botsch M.Example-Driven Deformations Based on Discrete Shells[C]//Computer Graphics Forum.Blackwell Publishing Ltd,2011,30(8):2246-2257.
[4]?stlund J,Varol A,Ngo D T,et al.Laplacian meshes for monocular 3D shape recovery[M]//Computer Vision-ECCV 2012.Springer Berlin Heidelberg,2012:412-425.
[5]Zeng D,Zheng D.Three-dimensional deformation in curl vector field[J].Journal of Zhejiang University SCIENCEC(S1869-1951),2012,13(8):565-572.
[6]Kohout J,Kellnhofer P,Martelli S.Fast deformation for modelling of musculoskeletal system[C]//Proceedings of the International Conference on Computer Graphics Theory and Applications and International Conference on Information Visualization Theory and Applications2012:16-25.
[7]Kraevoy V,Sheffer Aal.Modeling from contour drawings.Proceedings of the 6th Eurographics Symposium on Sketch-Based Interfaces and Modeling[C].New York,USA:ACM,2009:37-44.
[8]Xu K,Zheng H,Zhang H,et al.Photo-inspired model-driven 3D object modeling[C]//ACMTransactions on Graphics(TOG).Vancouver,Canada:ACM,2011:80.
[9]Entem E,Barthe L,Cani M P,et al.Modeling 3Danimals from a side-view sketch[J].Computers&Graphics(S0097-8493),2015,46(2):221-230.
[10]Cashman T J,Fitzgibbon A W.What shape are dolphins Building 3D morphable models from 2D Images[J].Pattern Analysis and Machine Intelligence,IEEETransactions on(S0162-8828),2013,35(1):232-244.
[11]Jacobson A,Baran I,Popovic J,et al.Bounded biharmonic weights for real-time deformation[J].ACMTrans.Graph(S0730-0301),2011,30(4):78.
[12]Borosán P,Jin M,DeCarlo D,et al.Rigmesh:automatic rigging for part-based shape modeling and deformation[J].ACM Transactions on Graphics(TOG)(S0730-0301),2012,31(6):198.
[13]Zhou S Z,Fu H,Liu L G,et al.Parametric reshaping of human bodies in images[C]//ACM Transactions on Graphics(TOG).USA:ACM,2010,29(4):126.
[14]Chen X B,Golovinskiy A,Funkhouser T.A benchmark for 3D mesh segmentation[C]//ACM Transactions on Graphics(TOG).USA:ACM,2009,28(3):73.
[15]于勇,王兆其,夏时洪,等.一种姿态无关的人体模型骨骼提取方法[J].计算机研究与发展,2008,45(7):1249-1258.Yu Yong,Wang Zhaoqi,Xia Shihong,et al.Pose Independent Joint Extraction from Scanned Human Body[J].Journal of Computer Research and Development,2008,45(7):1249-1258.