三維物體與場景模型的重建，在電腦繪圖、工業檢測與人機互動有許多重要的應用。在本文中，我們提出一個以幾何模型為基礎的三維重建方法，與傳統特徵為基礎的方法相比，此法最大特色在於物體表面只需有一組特徵對應點即可重建三維物體，特別適合表面顏色或紋路近似缺乏特徵的物體。

　　由於這些缺乏特徵的物體經常出現在原物料儲存場，它們通常由礦砂或礦石堆疊而成圓錐狀，因此我們選用圓錐作為基本幾何模型。首先，我們對影像進行背景去除並定義物體輪廓，根據這些輪廓計算圓錐幾何參數，然後以圓錐模型投影與影像上物體輪廓的重疊程度調整模型在空間中的大小、方位與位移；接著將調整後的模型切割成體素的集合作細部微調，以去除多餘的區域，接著計算物體可視表面並予以著色。若處理多物體的場景，則影像上常見物體相互遮蔽，使輪廓定義不明而影響重建，我們以初始模型反投影的對位解決此遮蔽問題，

　　本文引入模型概念重構三維物體，實驗顯示計算速度快且有令人滿意的效果，雖然商用雷射三維掃描器已能進行準確且高密度的場景重建，但是儀器昂貴且不普及。本方法僅需一般市售的數位相機，不僅成本低廉且取像方便，適合應用在原物料儲存場自動化的物料存取、環境評估和場景監控。

　Three-dimensional (3D) reconstruction from multiple images is a challenging problem in computer vision. However, it has many applications in industrial automa-tion, computer graphics, and human-computer interaction. In this thesis, we proposed a model-based 3D reconstruction method. Traditional methods usually fail to recon-struct the object without sufficient features. Fortunately, the proposed method only needs single feature point pair to build the 3D model; hence, it is suitable for objects with similar color or texture.

　Typical objects with very few features are usually seen in industrial applications, such as heap of stones/minerals stored in a large working field. The heap usually has a cone-like shape; hence, we use a parametric cone as the basic geometric model. At the beginning, we use background subtraction to detect the silhouette of the object for each image. The silhouettes are then used to estimate the parameters of the model. We adjust the size, position, and orientation of the estimated model by minimizing the overlapping ratio between the back-projections of model and the detected silhouettes in images. At last, the model is voxelized to facilitate redundant area removal, and surface coloring is accomplished with a visibility testing. For scenes with multiple objects, occlusion is frequently encountered and yields difficulties in the reconstruc-tion of individual object. This problem is formulated as a registration problem and solved by iterative closet point algorithm.

　Experimental results show that the proposed method is usually fast and the qual-ity of the recovered objects is satisfactory. Although three-dimensional (3D) scanners can build dense and accurate scene information nowadays, it is too expensive and can only work in limited environments. The proposed method only needs common digital cameras; therefore, it has great potential to be used as a low-cost solution in the auto-mation on mineral store and transportation.

[1] Y. Sato, M. Sugawara, Y. Tagawa, et al. “Automatic operation system for yard machines,” Kawasaki Steel Technical Report, no 13, Set, 1985.
[2] C. Chio, K. Lee, K. Shin, K. S. Hong, and H. Ahn. “Automatic Landing Method of a Reclaimer on the Stockpile,” IEEE Trans. Syst., Man, Cybern. C, vol. 29, no 1, pp. 308-314, Feb. 1999.
[3] Y. Xu, C. Xu, Y. Tian, S. Ma; M. Luo. “3D face image acquisition and reconstruc-tion system,” in Proc. IEEE Conf. Instrum. Meas. Technol., vol. 2, pp. 18-21, May. 1998.
[4] H. Kawasaki and R. Furukawa, “Entire model acquisition system using handheld 3D digitizer,” in Proc. 2nd Int. Conf. 3D Data Processing, Visualization, and Trans-mission., pp. 6–9, Sep. 2004.
[5] V. Sequeira, J.G. M. Goncalves, M. I. Ribeiro, “3D reconstruction of indoor envi-ronments,” in Proc. Int. Conf. Image Processing., vol. 1, pp. 16–19, Sep. 1996.
[6] S. Sullivan and J. Ponce “Automatic Model Construction and Pose Estimation From Photographs Using Triangular Splines,” IEEE Trans. Pattern Anal. Machine Intell., vol. 20, no. 10,pp. 1091–1097 Oct. 1998.
[7] C. I. Connolly and J. R. Stenstrom, “3D scene reconstruction from multiple inten-sity images,” in Proc. IEEE Workshop Interpretation of 3D Scenes., pp. 124–130, Austin, Tex., Nov. 1989.
[8] S. Srivastava and N. Ahuja, “Octree generation from object silhouettes in perspec-tive views,” Computer Vision, Graphics, and Image Processing., vol. 49, no. 1, pp. 68–84, Jan. 1990.
[9] P. Giblin and R. Weiss, “Reconstruction of surfaces from profiles,” in Proc. 1st Int. Conf. Computer Vision., pp. 136–144, London, June 1986.
[10] G.-Q. Wei and G. Hirzinger, “Parametric shape-from-shading by radial basis functions,” IEEE Trans. Pattern Anal. Machine Intell., vol. 19, no. 4, pp. 353–365, Apr. 1997.
[11] H. Maitre and W. Luo, “Using models to improve stereo reconstruction,” IEEE Trans. Pattern Anal. Machine Intell., vol. 14, no. 2, pp. 269–277, Feb. 1992.
[12] B. Moghaddam, J. Lee, H. Pfister and R. Machiraju, “Model-based 3D face cap-ture with shape-from-silhouette,” IEEE International Workshop on Analysis and Modeling of Faces and Gestures, pp. 20–27, 17 Oct. 2003.
[13] Y. I. Abdel-Aziz, H. M. Karara, “Direct linear transformation from comparator coordinates into object space coordinates in close-range photogrammetry,” in Proc. Proceedings of the Symposium on Close-Range Photogrammetry, 1971, pp. 1–18.
[14] R.Y. Tsai, “A versatile camera calibration technique for high-accuracy 3D ma-chine vision metrology using off-the-shelf TV cameras and lenses,” IEEE J. Robotics and Automation, vol. 3, no. 4, pp. 323–344, Aug. 1987.
[15] Z. Zhang, “Flexible camera calibration by viewing a plane from unknown orien-tations,” in Proc. Int. Conf. Computer Vision, Corfu, Greece, pp. 666–673, Sep. 1999.
[16] Z. Zhang, “A flexible new technique for camera calibration,” IEEE Trans. Pat-tern Anal. Machine Intell., vol. 22, no. 11, pp.1330–1334, Nov. 2000.
[17] http://www.vision.caltech.edu/bouguetj/calib_doc/
[18] J. Heikkila and O. Silven, “A four-step camera calibration procedure with im-plicit image correction,” in Porc. Conf. Computer Vision and Pattern Recognition, pp. 1106–1112, Jun. 1997.
[19] http://www.vision.caltech.edu/bouguetj/calib_doc/htmls/parameters.html
[20] R. C. Gonzalez and R. E. Woods, Digital Image Processing, 2nd ed., Prentice Hall, New Jersey, 2002
[21] C. Tomasi and T. Kanade, “Detection and tracking of point features,” Carnegie Mellon University, Tech. Rep. CMU-CS-91-132, April 1991.
[22] S. S. Rao, Engineering Optimization Theory and Practice, 3rd ed., Wiley-Interscience, New York, p357–p368, 1996.
[23] http://math.fullerton.edu/mathews/n2003/PowellMethodMod.html
[24] W. H. Press, B.P. Flannery, S.A. Teukolsky, and W.T. Vetterling, Numerical Recipes in C, 2nd ed., Cambridge University Press, Cambridge, U.K., 1992.
[25] http://www-cgrl.cs.mcgill.ca/~godfried/teaching/cg-projects/98/normand/main
.html
[26] C. Pluempitiwiriyawej, J. M. F. Moura, Y.-J. Lin Wu and C. Ho. “STACS: new active contour scheme for cardiac MR image segmentation,” IEEE Trans. Med. Imag., vol. 24, no. 5, pp. 593–603 May. 2005.
[27] S. D. Olabarriaga, J.-M. Rouet, M. Fradkin, M. Breeuwer and W. J. Niessen, “Segmentation of thrombus in abdominal aortic aneurysms from CTA with nonpara-metric statistical grey level appearance modeling,” IEEE Trans. Med. Imag., vol. 24, no. 4, pp. 477–485, Apr. 2005.
[28] C. Hernandez and F. Schmitt, “Multi-stereo 3D object reconstruction,” in Proc. 1st Int. Conf. 3D Data Processing Visualization and Transmission, Padova, Italy, pp. 19-21, Jun. 2002.
[29] A. Y. Mülayim, U. Yilmaz, and V. Atalay, “Silhouette-based 3-D model recon-struction from multiple images,” IEEE Trans. Syst., Man, Cybern. B, vol. 33, no. 4, pp. 582–591, Aug. 2003.
[30] A. O. Balan, “Voxel Carving and Coloring – Constructing a 3D model an object from 2D images,” unpublished.
[Online] Available: http://www.cs.brown.edu/people/alb/research.htm.
[31] S. M. Seitz and C. R. Dyer, “Photorealistic scene reconstruction by voxel
coloring,” in Proc. Int. J. Comput. Vis., vol. 35, no. 2, pp. 151–173, 1999.
[32] Y. Matsumoto, K. Fujimura, and T. Kitamura, “Shape-from-silhouette/stereo and its application to 3-d digitizer,” in Proc. Discrete Geometry Computer Imagery Conf., Lecture Notes Computer Science 1568, pp. 177–188, 1999.
[33] K. N. Kutulakos and S. M. Seitz, “A theory of shape by space carving,” in Proc. Int. J. Comput. Vis., vol. 38, no. 3, pp. 1999–218, 2000.
[34] W. B. Culbertson, T. Malzbender and G. Slabaugh, “Generalized voxel coloring,” in Proc. Int. Conf. Comput. Vis., pp. 100–115, Sep. 1999.
[35] L. Zhang and S. M. Seitz, “Image-based multiresolution modeling by surface deformation,” Robotics Institute Carnegie Mellon University, Tech. Rep. CMU-RI-TR-00-07, 2000.
[36] G. Slabaugh, B. Culbertson, T. Malzbender and R. Schafer, “A survey of methods for volumetric scene reconstruction from photographs”, Volume Graphics 2001, Proc. of the Joint IEEE TCVG and Eurographics Workshop, Springer Computer Science, pp. 81 - 100, 2001.
[37] P. J. Besl and N. D. McKay, “A method for registration of 3-D shapes,” IEEE Trans. Pattern Anal. Machine Intell., vol. 14, no. 2, pp. 239–256, 1992.
[38] B. K. P. Horn, “Closed-form solution of absolute orientation using unit quater-nion,” J.Opt. Soc. Amer. A, vol. 4, no. 4, pp. 629–642, Apr. 1987.
[39] O. D. Faugeras and M. Hebert, “The representation, recognition, and locating of 3-D object,” in Int. J. Robotic Res, vol. 5, no. 3, pp.27-52, Fall, 1986.
[40] J. R. Parker, Algorithms for image processing and computer vision, Wiley-Interscience, pp.176-219, 1999.