摘要
随着计算机技术的发展，人们已经不再满足于在计算机屏幕中看到简单的二维画面了。计算机图形学提供了一种在计算机上描述三维物体并且能够展现在用户面前的一种方法。但是，目前还只是停留在手工建模的基础之上，这就对用户提出了很高的要求-必须是专业人事才能做。而对于普通用户来说，对于模型精度要求不是很高，现在基本上是采用IBR(ImageBased Rendering)的方法。
本文讨论的就是一种通过多幅照片恢复物体三维模型的一种方法。从图像中恢复三维模型集成了图像处理、计算机视觉和计算机图形学的知识于一体，是一种学科门类跨度比较大的技术。但是在每个分支领域里面前人都已经做了很多的工作，基本上已经有了比较成熟的方法，这就使得整个系统在可行性上有了一定的保障。
首先介绍摄像机的透视投影模型，并讲述了普通摄像机标定的方法和两种平面模板的标定方法。然后给出了在我们系统中应用的自行设计的标定模板，最后以表格的形式给出了结果数据。
接下去讲述了体积相交求模型的方法。我们从普遍的体素定义引申开，介绍了我们非体素的表示方法。然后介绍了切割法的原理，并且推导出由三维实体相交转化为二维平面上线段相交的简化算法。最后给出了具体算法描述和实验结果。
第四章首先介绍了OpenGL的基本概念，然后介绍了利用平面的三角剖分来简化求解空间三角网格的生成算法。
最后一章给出系统的实验结果，并且讨论了目前系统存在的不足之处和改进方法。
With the development of computer technology, people are not satisfied with two dimensional images in computer screen. Computer Graphics (CG) provides a method of representing a three dimensional object in computer. However, these models are created manually. This work should be done by professional. Not a professional but a common user, mostly, what he need is not a high-precision model but a low-precision one. And now, many productions use the technology of Image Based Rendering (IBR).
In this thesis, we give the method of 3-D reconstruction over a set of pictures. This technology is based on Image Processing, Computer Vision and Computer Graphics. And we can find some good work in each subject so that we can use their results to construct our system.
Firstly, we give an introduction of the projective camera model. The next topic is camera calibration and we give an ordinary way to calibrate and two calibration method of using one model plane. Then we show the calibration model which is designed by us and finally give out our result of calibration.
Secondly, we explain the method of volume generation. From an ordinary definition of volxel, we give our way to express the volume. Then we introduce the principle of volume intersection. We also figure out that we can solve the intersection of two line segment in a 2-D image plane instead of calculate the intersection of two volume directly. Last we give the flow of our algorithm and the result of experiments.
In Chapter 4, we introduce the basic conception of OpenGL. And then we give a method to generate a local triangular mesh based on Delaunay triangulation technique in two dimensions.
In final chapter, we give out the result of our whole system. Then discuss the disadvantage of our system and give some new method to improve it.
引文
[1] S.E.Chen,"QuickTime VR-An Image-Based Approach to Virtual Environment Navigation",Proc.SIGGRAPH 95,p29-p38,1995
[2] Roger Y.Tsai,"An Efficient and Accurate Camera Calibration Technique for 3D MachineVision",Proceedings CVPR'86,Miami Beach,Florida,p364-p374,IEEE,June 1986
[3] Roger Y.Tsai,"A Versatile Camera Calibration Technique for High-Accuracy 3D MachineVision Metrology Using Off-the-Shelf TV Cameras and Lenses",IEEE Journal of Roboticsand Automation,vol.RA-3: p323-p341,August 1987
[4] Wei Qingchao,Zhou GuoQing,"On DLT Method for CCD Camera Calibration",Proceedings of ICSP' 96,p883-p884
[5] Zhengyou Zhang,"A Flexible New Technique for Camera Calibration",Technical ReportMSR-TR-98-71,Microsoft Research,December 1998
[6] Xiaoqiao Meng,Hua Li,Zhanyu Hu,"A New Easy Camera Calibration Technique Based onCircular Points",The Eleventh British Machine Vision Conference,11-14 September 2000
[7] O.D.Faugeras,G Toscani,"The Calibration Problem for Stereo",Proc.IEEE CVPR,p5-p20,1986
[8] Richard Hartley,Andrew Zisserman,"Multiple View Geometry in computer vision",Cambridge University Press,Cambridge,UK,2000．
[9] R.Koch,M.Pollefeys,B.Heigl,L.Van Gool,H.Niemann,"Calibration of Hand-heldCamera Sequences for Plenoptic Modeling",Proceedings ICCV'99,1999
[10] Deveemay,F.and Faugeras,O."From Projective to Euclidean Reconstruction",Proc.Computer vision and Pattern Recognition,p264-p269,1996
[11] Olivier Faugeras,Luc Robert,"What can two images tell us about a third one?",Int.J.of Computer Vision,18(1) :p5-p19,1996
[12] B.Curless,M.Levoy,"A Volumetric Method for Building Complex Models from RangeImages",SIGGRAPH'96 p303-p312,August 1996
[13] Andrew Fitzgibbon,Andrew Zisserman,"Automatic 3D Model Acquisition and Generationof New Images,from Video Sequences",http://citeseer.nj.nec.com/158432． html
[14] Andrew W.Fitzgibbon,Geoff Cross,Andrew Zisserman,"Automatic 3D Model Constructionfor Turn-Table Sequences",http://citeseer.nj.nec.com/119298． html
[15] M.Werman,A.Shashua,"The Study of 3D-from-2D using Elimination",http://citeseer.nj.nec.com/22771． html
[16] D.Liebowitz,A Criminisi,A.Zisserman,"Creating Architectural Models from Images",Proc.EuroGraphics,volume 18,p39-p50,Sep 1999
[17] R.Koch,"3-D Surface Reconstruction from Stereoscopic Image Sequences",http://citeseer.nj.nec.com/222545． html
[18] L.Falkenhagen,"3D Object-based Depth Estimation from Stereoscopic Image Sequences",
http://citeseer.nj.nec.com/falkenhagen95objectbased.html
[19] M. Pollefeys, R. Koch, M. Vergauwen, B. Deknuydt, L. V. Gool, "Three-dimensional scene reconstruction from images", http://citeseer.nj.nec.com/pollefeys00threedimensional.html
[20] M. Pollefeys, "Tutorial on 3D Modeling from Images", 26 June 2000, Dublin, Ireland, In conjunction with ECCV 2000
[21] P. Beardsley, P. Torr, A. Zisserman, "3D Model Acquisition from Extended Image Sequences", Proc. ECCV'96, Cambridge, U.K., 1996, p683-695
[22] C. E. Liedtke, R. Koch, "Modelling Of 3D Scenes From The Analysis Of Stereoscopic Image Sequences", http://citeseer.nj.nec.com/146920.html
[23] Hugues Hoppe, Tony DeRose, Tom Duchamp, John McDonald, Werner Stuetzle, "Surface Reconstruction from Unorganized Points", Computer Graphics (SIGGRAPH'92 Proceedings), Vol.26, No.2, July 1992, p71-p78
[24] Foley T A. Hagen, H. Nielson, "Visualizing and modeling unstructured data", The Visual Computer International Journal of Computer Graphics 1993.9(8): p439-p449
[25] David F. Waston, "Computing the n-Dimensional Delaunay Tessellation with Application to Voronoi Polytopes", Computer Journal, Vol.24, No.2, 1981, p167-p171
[26] N. Amenta, M. Bern, M. Kamvysselis, "A new Voronoi-Based Surface Reconstruction Algorithm", SIGGRAPH 1998
[27] Bowyer, A. "Computing Dirichlet Tessellations", The Computer Journal, Vol.24, No.2, 1981, p162-p166
[28] A. K. Cline, R. L. Renka, "Storage Efficient Method for Construction of a Thiessen Triangulation", Rockey Mountain Journal of Mathematics, Vol.14, No.1, 1984, p119-p139
[29] S. W. Sloan, "A Fast Algorithm for Constructing Delaunay Triangulations in the Plane", Advances in Engineering Software, 1987.9(1): p34-p55
[30] Xinguo Liu, Hujun Bao, PhengAnn Heng, TienTsin Wong, Qusheng Peng, "Constrained Fairing for Meshes", Computer Graphics Forum, March, 2001
[31] H. Hoppe, T. DeRose, T. Duchamp, J. McDonald, W. Stuetzle, "Mesh Optimization", Computer Graphics (SIGGRAPH'93 Proceedings), p19-p26, August 1993
[32] W. Press, B. Flannery, S. Teukolsky, W. Vetterling, "Numerical Recipes in C", Cambridge University Press, 1988
[33] 郑南宁，“计算机视觉与模式识别”，国防工业出版社
[34] 高文，陈熙霖，“计算机视觉——算法与系统原理”
[35] 朱心雄，“自由曲线曲面造型技术”，科学出版社
[36] 高满屯，曲仕茹，李西琴，“计算机视觉研究中的投影理论和方法”，西北工业大学出版社
[37] 王青，王融清，鲍虎军，彭群生，“散乱数据点的快速增量曲面重建算法”