基于单幅图片的唐代服饰三维重建
|
摘要
针对传统建模方法的局限性和对唐代服饰进行建模自身的特点,提出了一种将数字建模和物理建模相结合的基于单幅图片的唐代服饰的三维重建的方法。由于唐代历史悠久,遗留的下来的服装文物基本为二维水彩画,陶俑等,能从中提取的信息及其有限,因此利用Open CV的自动扫描技术和根据服装横截面近似椭圆的特性计算得出服装大致的数字模型,而数字模型只有服装的大致轮廓,无法呈现逼真的三维服饰,则再将数字模型的数据数据输入三维建模软件Maya中进行细节修改,并对其添加动力学特性,纹理贴图渲染。仿真结果显示三维重建出的唐代服饰模型,不仅逼真,而且效率高。
To overcome the limitations of traditional modeling methods,considering the specific feature of modeling of the Tang Dynasty costumes,this paper proposes a method based on single images of three-dimensional reconstruction of the Tang Dynasty costumes,combining digital modeling and physical modeling. Due to the long history of the Tang Dynasty,clothing artifacts left down are mainly the basic two-dimensional watercolor,pottery,etc.,the information that can be extracted is limited,so this,research uses using Open CV automatic scanning technology and the characteristics of the approximate oval cross-section calculated in accordance with clothing digital model of the clothing generally,and the digital model only shows the general outline of clothing,can not render realistic three-dimensional apparel,then the data are input in the Maya 3D modeling software and then carried out the digital model changes and add dynamics to its texture. Combining digital modeling and physical modeling improves the modeling efficiency,shortens modeling time,and reduces the distortion of the model as well.
To overcome the limitations of traditional modeling methods,considering the specific feature of modeling of the Tang Dynasty costumes,this paper proposes a method based on single images of three-dimensional reconstruction of the Tang Dynasty costumes,combining digital modeling and physical modeling. Due to the long history of the Tang Dynasty,clothing artifacts left down are mainly the basic two-dimensional watercolor,pottery,etc.,the information that can be extracted is limited,so this,research uses using Open CV automatic scanning technology and the characteristics of the approximate oval cross-section calculated in accordance with clothing digital model of the clothing generally,and the digital model only shows the general outline of clothing,can not render realistic three-dimensional apparel,then the data are input in the Maya 3D modeling software and then carried out the digital model changes and add dynamics to its texture. Combining digital modeling and physical modeling improves the modeling efficiency,shortens modeling time,and reduces the distortion of the model as well.
引文
[1]孙宇阳.基于单幅图像的三维重建技术综述[J].北方工业大学学报,2011,23(1):9-13.
[2]蔡钦涛.基于图像的三维重建技术研究[D].杭州:浙江大学,2004.
[3]朱庆生,罗大江,葛亮,等.基于多幅图像的三维重建[J].计算机工程与设计,2010,31(10):2351-2353.
[4]张娟.基于多视图几何三维重建[D].南京:南京航空大学,2012.
[5]赵璐璐,耿国华,王小凤,等.基于未标定多幅图的三维重建算法[J].计算机应用,2012,32(10):2802-2805.
[6]何理想.基于SFS方法的单幅图像的三维原型重建的研究[D].沈阳:沈阳工业大学,2009.
[7]高欣健,张旭东,高亚捷,等.基于单幅灰度图像的快速三维重建方法研究[EB/OL].http://www.cnki.net/kcms/detail/11.2187.TH.20130813.1354.016.html,2013-08-13.
[8]彭如.基于自阴影处理的SFS三维重建算法研究[D].南昌:南昌航空大学,2011.
[9]仇开金.基于单幅图像明暗恢复形状算法设计[J].计算机与现代化,2010,6(6):58-59.
[10]吴凤和,张晓峰,施法中.基于单幅图像数据的三维重构方法研究[J].中国机械工程,2007,18(17):2071-2075.
[11]高月芳,罗飞,曹建忠.由单幅二维灰度图像重构物体表面形状[J].计算机应用研究,2007,24(8):195-197.
[12]HSIEH Jun-wei,LIAO MARK Hong-yuan,KO Mingtat,et al.Wavelet-based shape from shading[J].Graphical Models and Image Processing,1995,57(4):1077-3169.
[13]ELHABIAN S,MOSTAFA E,RARA H,et al.NonLamebrain model-based facial shape recovery from single image under unknown general illumination[C].Conference on Computer and Robot Vision,Toronto:IEEE,2012:252-259.
[14]秦小文,温志芳,乔维维.基于Open CV的图像处理[J].电子测试,2011,7(7):39-41.
[15]黄辉.中国历代服制服式[M].南昌:江西美术出版社,2011:114-164.
[2]蔡钦涛.基于图像的三维重建技术研究[D].杭州:浙江大学,2004.
[3]朱庆生,罗大江,葛亮,等.基于多幅图像的三维重建[J].计算机工程与设计,2010,31(10):2351-2353.
[4]张娟.基于多视图几何三维重建[D].南京:南京航空大学,2012.
[5]赵璐璐,耿国华,王小凤,等.基于未标定多幅图的三维重建算法[J].计算机应用,2012,32(10):2802-2805.
[6]何理想.基于SFS方法的单幅图像的三维原型重建的研究[D].沈阳:沈阳工业大学,2009.
[7]高欣健,张旭东,高亚捷,等.基于单幅灰度图像的快速三维重建方法研究[EB/OL].http://www.cnki.net/kcms/detail/11.2187.TH.20130813.1354.016.html,2013-08-13.
[8]彭如.基于自阴影处理的SFS三维重建算法研究[D].南昌:南昌航空大学,2011.
[9]仇开金.基于单幅图像明暗恢复形状算法设计[J].计算机与现代化,2010,6(6):58-59.
[10]吴凤和,张晓峰,施法中.基于单幅图像数据的三维重构方法研究[J].中国机械工程,2007,18(17):2071-2075.
[11]高月芳,罗飞,曹建忠.由单幅二维灰度图像重构物体表面形状[J].计算机应用研究,2007,24(8):195-197.
[12]HSIEH Jun-wei,LIAO MARK Hong-yuan,KO Mingtat,et al.Wavelet-based shape from shading[J].Graphical Models and Image Processing,1995,57(4):1077-3169.
[13]ELHABIAN S,MOSTAFA E,RARA H,et al.NonLamebrain model-based facial shape recovery from single image under unknown general illumination[C].Conference on Computer and Robot Vision,Toronto:IEEE,2012:252-259.
[14]秦小文,温志芳,乔维维.基于Open CV的图像处理[J].电子测试,2011,7(7):39-41.
[15]黄辉.中国历代服制服式[M].南昌:江西美术出版社,2011:114-164.