摘要
三维表面重建是计算机视觉领域的重要研究方向,其目的是从二维信息中恢复重建对象的三维特征,建立对象的几何模型。本文深入研究了基于目标物体的轮廓信息进行三维表面重建的算法,设计并完成了一个多视角实时三维重建系统。
本文使用固定的多台摄像机,从不同视角采集自由运动目标的图像,通过自适应的可见外壳生成算法,实现运动物体的实时三维重建,目的是提高重建对象的精度和三维重建的可靠性与灵活性。研究内容在三维建模、三维展示、虚拟现实等方向有着广泛的应用前景。
论文首先介绍了三维重建领域的研究背景,在此基础上,结合课题研究的实际需要,提出了论文的主要研究内容:图像预处理、摄像机标定以及基于自适应距离场的实时三维重建。图像预处理的主要目的是从图像中滤除噪声,分离出对后续处理有用的信息。我们采用背景减除和数学形态学方法实现运动目标与图像背景与的分离。在摄像机标定方法研究中,我们采用介于传统标定方法和自标定方法之间的张氏平面标定法,准确地获得了摄像机的内外参数。在重建对象三维信息的研究中,根据从轮廓恢复形状的基本原理,结合了自适应距离场,对运动目标进行实时三维重建。最后通过分析系统误差和瓶颈,提出了改进方案。
总之,本文不仅从理论上阐述了基于图像的实时三维重建方法,而且也给出了具体的系统实现方案,其研究工作具有理论意义和应用价值。
3D surface reconstruction is an unsolved and important task in computer vision. It aims at restoring three-dimensional features from two-dimensional information and constructing the geometrical model of objects. This thesis focuses on algorithms of the three-dimensional surface reconstruction from silhouette information of objects, designs and develops a multi-view real-time 3D reconstruction system.
In this thesis, multiple fixed cameras are used for collecting images of moving objects from different perspectives. By generating visual hulls with an adaptive algorithm, the surfaces of moving objects are reconstructed in real time. Our aim is to improve the accuracy, reliability and flexibility of three-dimensional reconstruction. The research has a broad prospect in 3D modeling, representation and virtual reality.
First of all, this thesis introduces the research background of the area of 3D reconstruction. On this basis, with the actual needs of our system, we propose the main elements of this thesis: image pre-processing, calibration and adaptive distance field based real-time 3D reconstruction. The purpose of image pre-processing is to filter out noise from the image and to isolate useful information for follow-up processes. We adopt background subtraction and mathematical morphology algorithms to separate moving objects from background. In the research of calibration algorithm, by using Zhang's plane calibration which is between the traditional algorithm and self-calibration algorithm, we can acquire the precise parameters of cameras. In the research of reconstruction, based on the principles of Shape from Silhouette, combined with Adaptive Distance Fields, we implement the real-time reconstruction of moving objects. Finally, we analysis the system errors and bottlenecks, and introduce the scheme of future improvement.
In conclusion, this thesis not only theoretically expounds an image-based real-time reconstruction algorithm, but also gives an implementation of the specific system. This research work has value of both theory and application.
引文
[1] P.E. Debevec, C.J. Taylor, and J. Malik, Modeling and rendering architecture from photographs: a hybrid geometry-and image-based approach, ACM Press New York, NY, USA, 1996.
[2] D. Marr, Vision: A Computational Investigation into the Human Representation and Processing of Visual Information, Henry Holt and Co., Inc. New York, NY, USA, 1982.
[3] B.K.P. Horn and M.J. Brooks, Shape from shading, MIT Press Cambridge, Mass, 1989.
[4] J.J. Gibson, The Ecological Approach to Visual Perception, Lawrence Erlbaum Associates, 1986.
[5] M. Montani, C. Rocchini, and K.O.T. Persson, "Marching Intersections; An Efficient Approach to Shape-from-Silhouette," Vision, Modeling, and Visualization 2002, 2002.
[6] G.K.M. Cheung, S. Baker, and T. Kanade, "Visual hull alignment and refinement across time: a 3D reconstruction algorithm combining shape-from-silhouette with stereo," Computer Vision and Pattern Recognition, 2003. Proceedings. 2003 IEEE Computer Society Conference on, 2003,2.
[7] C. Tomasi and T. Kanade, "Shape and Motion from Image Streams under Orthography: a Factorization Method," InternationalJournal of Computer Vision, 1992,9: 137-154.
[8] O. Faugeras, Three-Dimensional Computer Vision: A Geometric Viewpoint, Mit Press, 1993.
[9] B.H. Kim and R.H. Park, "Shape from Shading and Photometric Stereo Using Surface Approximation by Legendre Polynomials," Computer Vision and Image Understanding, 1997,66: 255-270
[10] K.M. Lee and C.C.J. Kuo, "Surface reconstruction from photometric stereo images," Journal of the Optical Society of America A: Optics, Image Science, and Vision, 1993, 10: 855-868.
[11] J.W. Tukey, Exploratory data analysis, Reading, Mass.: Addison-Wesley, 1977.
[12] X. Wang and D. Wang, "On the max/median filter," Acoustics, Speech, and Signal Processing [see also IEEE Transactions on Signal Processing], IEEE Transactions on, 1990, 38: 1473-1475.
[13] G.R. Arce and R.E. Foster, "Detail-preserving ranked-order based filters for image processing," Acoustics, Speech, and Signal Processing [see also IEEE Transactions on Signal Processing], IEEE Transactions on,1989,37:83-98.
[14] I. Haritaoglu, D. Harwood, and L.S. Davis, "W4: Real-Time Surveillance of People and Their Activities,"2000.
[15] S.J. McKenna et al., "Tracking groups of people," Computer Vision and Image Understanding, 2000, 80:42-56.
[16] M. Kilger, "A shadow handler in a video-based real-time traffic monitoringsystem," Applications of Computer Vision, Proceedings, 1992., IEEE Workshop on: 11-18.
[17] C. Stauffer and W.E.L. Grimson, "Adaptive background mixture models for real-time tracking," Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 1999, 2: 246-252.
[18] 阮秋琦, 数字图像处理学M], 北京: 电子工业出版社, 2001.1:429-466.
[19] W. Faig, "Calibration of close-range photogrammetry systems: Mathematical formulation," Photogrammetric Engineering and Remote Sensing, 1975, 41: 1479-1486.
[20] M. Apollonio et al., "Initial Results from the CHOOZ Long Baseline Reactor Neutrino Oscillation Experiment," Arxiv preprint hep-ex/9711002,1997.
[21] D.B. Gennery, "Stereo-Camera Calibration.," 1979.
[22] L. Paquette et al., "A new camera calibration method for robotic vision," 1990: 656-663.
[23] R.Y. Tsai, "An efficient and accurate camera calibration technique for 3D machine vision," Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, 1986, 374.
[24] R.Y. Tsai, "A Versatile Camera Calibration Technique for High-Accuracy 3D Machine Vision Metrology Using Offthe-Shelf TV Cameras and Lenses," IEEE Journal of Robotics and Automation, 1987, 3: 323-344.
[25] O.D. Faugeras, Q.T. Luong, and S.J. Maybank, "Camera self-calibration: theory and experiments," European Conference on Computer Vision, 1992, 588: 321-334.
[26] R.I. Hartley, "Self-calibration from multiple views with a rotating camera," Proceedings of the third European conference on Computer vision (vol. 1) table of contents, 1994: 471-478.
[27] S. De Ma, "A self-calibration technique for active vision systems," Robotics and Automation, IEEE Transactions on, 1996, 12: 114-120.
[28] Z. Zhang, "Flexible camera calibration by viewing a plane from unknown orientations," International Conference on Computer Vision, 1999,1: 666-673.
[29] Z. Zhang, "A Flexible New Technique for Camera Calibration," IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22: 1330-1334.
[30] A. Laurentini, "The Visual Hull Concept for Silhouette-Based Image Understanding," IEEE Trans. Pattern Anal. Mach. Intell., 1994,16: 150-162.
[31] W. Matusik, C. Buehler, and L. McMillan, "Polyhedral Visual Hulls for Real-Time Rendering," Proceedings of the 12th Eurographics Workshop on Rendering Techniques, Springer-Verlag, 2001: 115-126.
[32] A. Kaufman, D. Cohen, and R. Yagel, "Volume graphics," Computer, 1993,26: 51-64.
[33] A. Rosenfeld and J.L. Pfaltz, "Sequential Operations in Digital Picture Processing," J. ACM, 1966, 13:471-494.
[34] Frisken, S., Perry, R., Rockwood, A., et al. Adaptively Sampled Distance Fields: A General Representation of Shape for Computer Graphics. In: Proceedings of SIGGRAPH '00. New York: ACM Press, 2000. 249-254.
[35] H. Samet, The Design and Analysis of Spatial Data Structures, Addison-Wesley, 1989.
[36] W.E. Lorensen and H.E. Cline, "Marching cubes: A high resolution 3D surface construction algorithm,"Proceedings of the 14th annual conference on Computer graphics and interactive techniques, 1987: 163-169.
[37] R.N. Perry and S.F. Frisken, "Kizamu: a system for sculpting digital characters," Proceedings of the 28th annual conference on Computer graphics and interactive techniques, 2001: 47-56.
[38] S.F.F. Gibson, "Using distance maps for accurate surface representation in sampled volumes," Proceedings of the 1998 IEEE symposium on Volume visualization, 1998: 23-30.
[39] Intel(?) VTune~(TM) Performance Analyzer; http://www.intel.com/cd/software/products/asmo-na/eng/239144.htm