双目立体测量中的人脸数据自拼合方法

刘兆宁; 张丽艳; 石春琴

双目立体测量中的人脸数据自拼合方法

1.
南京航空航天大学机电学院, 南京 210016

基金项目:

江苏省科技支撑计划项目(BK2008136)资助

详细信息

作者简介:
刘兆宁(1985-)，硕士研究生，研究方向为逆向工程和非接触三维几何测量，liuzhaoning513@163.com;张丽艳(联系人)，教授，博士生导师，zhangly@nuaa.edu.cn

刘兆宁(1985-)，硕士研究生，研究方向为逆向工程和非接触三维几何测量，liuzhaoning513@163.com;张丽艳(联系人)，教授，博士生导师，zhangly@nuaa.edu.cn

计量
- 文章访问数: 207
- HTML全文浏览量: 42
- PDF下载量: 4
- 被引次数: 0
出版历程
- 收稿日期: 2010-10-16
- 刊出日期: 2015-06-10

Automatic Registration of Facial Data in Binocular Stereo Measurement

1.
College of Mechanical and Electronic Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016

摘要

摘要: 提出一种在双目立体视觉测量中三维人脸数据的自拼合方法。该方法利用从不同视角下摄取的人脸图像,分别对这些图像进行特征点提取,并建立图像间同名特征点对的稳定匹配,从而获取人脸上相同位置在不同视角下的对应关系,然后利用五点算法求取不同视角下测量系统的相对位置和姿态,最终实现对多视角下三维人脸测量数据的自动拼合。提出的方法无需借助辅助标记点或者手动选择特征点,实验结果验证了该方法的有效性和鲁棒性。
- 双目立体测量 /
- 五点算法 /
- 自动拼合
Abstract: An automatic registration algorithm of facial data in binocular stereo measurement is proposed.Firstly,the point features are extracted from the facial images in different views,and then the corresponding point matches are steadily established between the images.Secondly,the five-point algorithm is employed to compute the relative pose between the two views to realize the auto-registration of 3D facial data.It is unnecessary to use markers for assistance or select the features by hand.Experiments demonstrate the validity and robustness of the proposed algorithm.
- binocular stereo measurement /
- the five-point algorithm /
- automatic registration

HTML全文

参考文献(14)

[1]	马颂德,张正友.计算机视觉-计算理论与算法基础[M].北京:科学出版社,1998
[2]	龙玺,钟约先,李仁举等.结构光三维扫描测量的三维拼接技术[J].清华大学学报,2002,42(4):477～480
[3]	Ferrario V F,Sforza C,Poggio C E,et al.Facial three-dimen-sional morphometry[J].American Journal of Orthodonticsand Dentofacial Orthopedics,1996,109(1):86～93
[4]	Pighin F,Hecker J,Lischinskiy D,Szeliski R,Salesin D H.Synthesizing realistic facial expressions from photographs[A].In:Proceedings of the SIGGRAPH98[C],Orlando,FL,USA,1998:75～84
[5]	邓雁萍,李保洲,刘志敏.人脸三维测量中的立体视觉[J].上海交通大学学报,1998,32(9):97～99
[6]	焦婷,张富强,孙健.应用三维激光扫描仪采集及重建头面部软组织的研究[J].上海口腔医学,2005,14(5):463～466
[7]	Liu Z H,Zhang Z Y,et al.Rapid Modeling of AnimatedFaces From Video[R].Microsoft Technical Report MSR-TR,2000,11:475～476
[8]	Lowe D G.Distinctive image features from scale-invariant key-points[J].International Journal of Computer Vision,2004,60(2)
[9]	石春琴,张丽艳,韦虎.随机纹理图像的特征匹配鲁棒算法[J].南京航空航天大学学报,2010,42(4):440～446
[10]	Nister D.An efficient solution to the five-point relative pose prob-lem[J].IEEE Transactions on Pattern Analysis and Ma-chine Intelligence,2004,26(6):756～770
[11]	Fischler M A,Bolles R C.Random sample consensus:a para-digm for model fitting with application to image analysis and auto-mated cartography[J].Communications of the ACM,1981,24(6):381～395
[12]	Zhang H,Zhang L Y,Wang H T,et al.Surface measurementbased on instantaneous random illumination[J].Chinese Jour-nal of Aeronautics,2009,22(3):316～324
[13]	张辉,张丽艳,陈江等.基于平面模板自由拍摄的双目立体测量系统的现场标定[J].航空学报,2007,28(3):695～701
[14]	Aiger D,Niloy M J,Cohen-Or D.4-points congruent sets for ro-bust pairwise surface registration[J].ACM Transactions onGraphics,2008,27(3):1～10