基于单幅编码图像的三维重构系统集成
|
摘要
本论文主要针对机器视觉中彩色伪随机序列编码与灵活的预标定三维重构技术，研究了基于单幅编码图像的三维欧氏重构系统集成。将一个内外参数变化可调的CCD相机摄取带编码信息的三维场景的图像到最后实现三维欧氏重构集成到一起：先对CCD相机采集的编码图像进行必要的预处理，接着对其进行特征点检测与匹配，结合预先标定的LCD投影仪参数，利用层次化的三维欧氏重构理论，即可根据单幅图像恢复该三维场景的三维坐标数据、重构三维场景的三维形貌。
所研究的内容主要包括以下三方面：在图像特征点检测与匹配方面，主要利用彩色伪随机序列编码三维场景特征化技术，根据编码设计的原理来确定特征点检测和匹配的方案。用LCD投影仪将彩色伪随机序列编码投射到三维场景表面，用CCD相机采集到图像，并对其进行必要的预处理，采用Harris角点检测方法提取编码特征点，然后利用伪随机序列编码的窗口特性，解决三维表面重构时匹配点识别难题；在三维重构的方法上，把彩色编码结构光投影仪作为一个逆向的相机予以考虑，使整个视觉系统就可以视为含有两个相机的典型的双目立体视觉系统，首先对相机的镜头畸变模型进行标定，保证后续投影仪预标定的准确性，但是在进行重构实验中相机的内外参数可以根据场景需要任意改变；在三维重构的算法上，采用层次化思想，在射影层次、仿射层次、尺度层次以及欧氏层次上建立视觉系统的成像模型，逐步实现三维重构，最终实现单幅图像的三维欧氏重构。
This paper mainly focuses on the integration of 3D reconstruction system from single encoded image based on the multicolor code of pseudo-random array and flexible pre-calibrated 3D reconstruction technology, which integrates from an encoded image to 3D Euclidean reconstruction at last. It can recover the 3D data and reconstruct the 3D shape of scene by a LCD projector whose parameters need to be pre-calibrated and a CCD camera whose parameters are allowed to change with the varying scene.
The integrated system mainly contains three parts as follows: The method of interest point detection and correspondence is decided by the multicolor code of pseudo-random array principle which made the 3D scene characterization. The multicolor code of pseudo-random array is projected to the surface of 3D scene. Then do some pre-process to the images captured by the CCD camera, and detect the interest point with the improved Harris corner detector. Each interest point on scene surface can be identified exclusively and the problem of interest point correspondence between images can be solved according to the window property of pseudo-random array. The vision system of 3D reconstruction is equal to typical stereo vision system because the projector is regarded as inverse camera, in which the lens distortion of camera must be pre-calibrated to guarantee the precision of projector pre-calibration, and the intrinsic and extrinsic parameters of camera can be modified according to the changing scene. The reconstruction algorithm is based on stratification idea. The imaging models of camera and projector are founded on four strata as Projective stratum, Affine stratum, Metric stratum and Euclidean stratum. At last, the idea of 3D Euclidean reconstruction of single image can be carried out.
所研究的内容主要包括以下三方面：在图像特征点检测与匹配方面，主要利用彩色伪随机序列编码三维场景特征化技术，根据编码设计的原理来确定特征点检测和匹配的方案。用LCD投影仪将彩色伪随机序列编码投射到三维场景表面，用CCD相机采集到图像，并对其进行必要的预处理，采用Harris角点检测方法提取编码特征点，然后利用伪随机序列编码的窗口特性，解决三维表面重构时匹配点识别难题；在三维重构的方法上，把彩色编码结构光投影仪作为一个逆向的相机予以考虑，使整个视觉系统就可以视为含有两个相机的典型的双目立体视觉系统，首先对相机的镜头畸变模型进行标定，保证后续投影仪预标定的准确性，但是在进行重构实验中相机的内外参数可以根据场景需要任意改变；在三维重构的算法上，采用层次化思想，在射影层次、仿射层次、尺度层次以及欧氏层次上建立视觉系统的成像模型，逐步实现三维重构，最终实现单幅图像的三维欧氏重构。
This paper mainly focuses on the integration of 3D reconstruction system from single encoded image based on the multicolor code of pseudo-random array and flexible pre-calibrated 3D reconstruction technology, which integrates from an encoded image to 3D Euclidean reconstruction at last. It can recover the 3D data and reconstruct the 3D shape of scene by a LCD projector whose parameters need to be pre-calibrated and a CCD camera whose parameters are allowed to change with the varying scene.
The integrated system mainly contains three parts as follows: The method of interest point detection and correspondence is decided by the multicolor code of pseudo-random array principle which made the 3D scene characterization. The multicolor code of pseudo-random array is projected to the surface of 3D scene. Then do some pre-process to the images captured by the CCD camera, and detect the interest point with the improved Harris corner detector. Each interest point on scene surface can be identified exclusively and the problem of interest point correspondence between images can be solved according to the window property of pseudo-random array. The vision system of 3D reconstruction is equal to typical stereo vision system because the projector is regarded as inverse camera, in which the lens distortion of camera must be pre-calibrated to guarantee the precision of projector pre-calibration, and the intrinsic and extrinsic parameters of camera can be modified according to the changing scene. The reconstruction algorithm is based on stratification idea. The imaging models of camera and projector are founded on four strata as Projective stratum, Affine stratum, Metric stratum and Euclidean stratum. At last, the idea of 3D Euclidean reconstruction of single image can be carried out.
引文
[1] YF Li, Zhang B. A Method for 3D measurement and reconstruction for active vision. Measurement Science and Technology, Vol.15, No. 11, Nov. 2004: 2224-2232.
[2] F. Jessie, M.Williams, N.J.A Fsloane, Pseudo-Random Sequences and Arrays[J], Proceedings of the IEEE. Vol.64(12), 1976: 1715-1728.
[3] A. Fusiello. Uncalibrated Euclidean reconstruction: a review. Image Vision Comput. 18(6-7), 2000: 555-563.
[4] S. J. Maybank, O. D. Faugeras. A theory of self-calibration of a moving camera. International Journal of Computer Vision, Vol.8, No.2,1992: 123-151.
[5] O. Faugeras, Q.T. Luong, S. Maybank. Camera self-calibration: Theory and experiments. Proceedings of the 2 European Conference Computer Vision, Italy, 1992: 321-334.
[6] R. Hartley. Euclidean reconstruction and invariants from multiple images. IEEE Transactions on Pattern Analysis and Machine Intelligence, 16(10), 1994: 1036-1041.
[7] B. Triggs. Auto-calibration and the absolute quadric. Proceedings of Computer Vision and Pattern Recognition, 1997: 609-614.
[8] M. Pollefeys, V. Gool, A. Oosterlinck, The modulus constraint: A new constraint for self-calibration. Proceedings of Internaltional Conference of Pattern Recongnition, Vienna, 1996: 349-353.
[9] A. Heyden, K. Astrom, Euclidean reconstruction from image sequences with varying and unknown focal length and principal point. Proc. of IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Juan, Puerto Rico, 1997: 438-443.
[10] M. Pollefeys, R. Koch and L. V. Gool. Self-calibration and metric reconstruction inspite of varying and unknown intrinsic camera parameters. International Journal of Computer Vision, Vol.32, No.1, pp.7-25,1999.
[11] Gang Liu. Studies on Geometry and Texture Reconstruction from Multiple Images. A Doctoral Dissertation of Zhejiang University, 2004.
[12] D. Fofi, J. Salvi, E. M. Mouaddib. Euclidean reconstruction by means of an uncalibrated structured light sensor. 4th. Ibero-American Symposium on Pattern Recognition (SIARP'2000), Lisboa (Portugal), 2000: 159-170.
[13] R. I. Hartley. Euclidean reconstruction from uncalibrated views, Applications of Invariance in Computer Vision. Lecture Notes in Computer Science, 852, Springer, Berlin, 1993: 237-256.
[14] D. Fofi, J. Salvi, E. Mouaddib, Uncalibrated Reconstruction: An Adaptation to Structured Light Vision, Pattern Recognition, 36(7), Juillet, 2003: 1631-1644.
[15] O. D. Faugeras and G. Toscani. The Calibration Problem for Stereo. CVPR86, 1986: 15～20.
[16] C. Chen and Y. F. Zheng. Passive and active stereo vision for smooth surface detection of deformed plates. IEEE Trans. on Industrial Electronics, No.42, No.3, 1995: 300-306.
[17] 陈乐，吕文阁，丁少华．角点检测技术研究进展．自动化技术与应用，2005，24(5)：1-4．
[18] Harris C G, Stephens M J, A Combined Comer and Edge Detector, Proceedings Fourth Alvey Vision Conference, Manchester, 1988, 147-151
[19] Ajinkya Chavan, Sreela Sasi, Vision-based Real-Time Speed Tracking, IEEE, 1-3 September, 2005. Faro, Portugal: 16-21.
[20] 高春鸣，兰秋军．应用数字图像增强技术处理医学影像的研究．计算机工程，2001．5
[21] 卢荣胜．视觉准直在线测量技术研究．天津大学精密仪器与光电子工程学院．2000．
[22] 徐奕，周军，周源华．立体视觉匹配技术．计算机工程与应用，2003(15)：1-5．
[23] 边缘和灰度信息结合的SAR图像配准方法研究[D]．北京：中国科学院电子学研究所，2003．
[24] T. Y. Tsai. An efficient and accurate camera calibration technique for 3D machine vision. IEEE Proceedings of Conf. on Computer Vision and Pattern Recognition[C], 1986: 364-374.
[25] T. S. Huang, A. N. Netravali. Motion and structure from feature correspondence: a review, Proceedings of the IEEE, Vol. 82, Vol. 2, 1994: 252-268.
[26] J. Weng, P. Cohen and M. Hemiou. Camera calibration with distortion models and accuracy evaluation. IEEE Trans. Patt. Anal. Machine Intell., 14(10), 1992: 965-980.
[27] 毛剑飞，邹细勇，诸静．改进的平面模板两步法标定摄像机．中国图象图形学报，2004，9(7)：846-852．
[28] Luca Lucchese, Geometric calibration of digital cameras through multi-view rectification. Image and Vision Computing 23 (2005): 517-539.
[29] Joaquim Salvi, Xavier Armangue, Joan Batlle. A comparative review of camera calibrating methods with accuracy evaluation. Pattern Recognition, 2002, 35(1):1617-1635.
[30] 张勇斌．基于单幅编码图像三维场景不标定欧氏重构技术．合肥工业大学博士论文．
[31] 戴丽，顾建刚．最小二乘圆法及其在原木定心中的应用．研究设计，2002(2)．
[32] J. G. Semple and G. T. Kneebone. Algebraic projective geometry[M]. Oxford University Press, New York, 1998: 23-49.
[33] D. Q. Huynh, R. A. Owens, Hartmann, PE. Calibrating a structured light stripe system: a novel approach. International Journal of Computer Vision, Vol.33, No. 1, 1999: 73-86.
[34] O. Faugeras. Stratification of three-dimensional vision: projective, affine, and metric Representations. Journal of the Optical Society of America A, Vol. 12, No. 3, 1994: 465-484.
[35] C. Edberg, A. Ericsson, 3D reconstruction from uncalibrated images[D], Lund Institute of Technology, Lund University, Sweden, December, 2000: 19-79.
[36] Z. Y. Zhang, Determining the Epipolar Geometry and its Uncertainty: A Review. The International Journal of Computer Vision, 27(2): 161-195, Also Research Report No. 2927, INRIA Sophia-Antipolis, March 1998.
[37] M. Pollefeys, L. Van Gool. A stratified approach tometrie self-calibration. Proceedings of the Conference on Computer Vision and Pattern Recognition, Puerto Rico, USA IEEE Computer Society Press, 1997: 407-412.
[38] Ame Henrichsen. 3D reconstruction and camera calibration from2 D images. University of Cape Town, 2000:7-24, 73-88.
[39] R. I. Hartley, In defence of the 8-point algorithm. IEEE Proc. of Fifth International Conference on Computer Vision, 1995:1064-1070.
[40] M. Pollefeys. Self-calibration and metric 3D reconstruction from uncalibrated image sequences. Ph D thesis, ESAT-PSI, K. U. Leuven, 1999.
[41] 施晓红，周佳．精通GUI图形界面编程．北京大学出版社，2003．
[2] F. Jessie, M.Williams, N.J.A Fsloane, Pseudo-Random Sequences and Arrays[J], Proceedings of the IEEE. Vol.64(12), 1976: 1715-1728.
[3] A. Fusiello. Uncalibrated Euclidean reconstruction: a review. Image Vision Comput. 18(6-7), 2000: 555-563.
[4] S. J. Maybank, O. D. Faugeras. A theory of self-calibration of a moving camera. International Journal of Computer Vision, Vol.8, No.2,1992: 123-151.
[5] O. Faugeras, Q.T. Luong, S. Maybank. Camera self-calibration: Theory and experiments. Proceedings of the 2 European Conference Computer Vision, Italy, 1992: 321-334.
[6] R. Hartley. Euclidean reconstruction and invariants from multiple images. IEEE Transactions on Pattern Analysis and Machine Intelligence, 16(10), 1994: 1036-1041.
[7] B. Triggs. Auto-calibration and the absolute quadric. Proceedings of Computer Vision and Pattern Recognition, 1997: 609-614.
[8] M. Pollefeys, V. Gool, A. Oosterlinck, The modulus constraint: A new constraint for self-calibration. Proceedings of Internaltional Conference of Pattern Recongnition, Vienna, 1996: 349-353.
[9] A. Heyden, K. Astrom, Euclidean reconstruction from image sequences with varying and unknown focal length and principal point. Proc. of IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Juan, Puerto Rico, 1997: 438-443.
[10] M. Pollefeys, R. Koch and L. V. Gool. Self-calibration and metric reconstruction inspite of varying and unknown intrinsic camera parameters. International Journal of Computer Vision, Vol.32, No.1, pp.7-25,1999.
[11] Gang Liu. Studies on Geometry and Texture Reconstruction from Multiple Images. A Doctoral Dissertation of Zhejiang University, 2004.
[12] D. Fofi, J. Salvi, E. M. Mouaddib. Euclidean reconstruction by means of an uncalibrated structured light sensor. 4th. Ibero-American Symposium on Pattern Recognition (SIARP'2000), Lisboa (Portugal), 2000: 159-170.
[13] R. I. Hartley. Euclidean reconstruction from uncalibrated views, Applications of Invariance in Computer Vision. Lecture Notes in Computer Science, 852, Springer, Berlin, 1993: 237-256.
[14] D. Fofi, J. Salvi, E. Mouaddib, Uncalibrated Reconstruction: An Adaptation to Structured Light Vision, Pattern Recognition, 36(7), Juillet, 2003: 1631-1644.
[15] O. D. Faugeras and G. Toscani. The Calibration Problem for Stereo. CVPR86, 1986: 15～20.
[16] C. Chen and Y. F. Zheng. Passive and active stereo vision for smooth surface detection of deformed plates. IEEE Trans. on Industrial Electronics, No.42, No.3, 1995: 300-306.
[17] 陈乐，吕文阁，丁少华．角点检测技术研究进展．自动化技术与应用，2005，24(5)：1-4．
[18] Harris C G, Stephens M J, A Combined Comer and Edge Detector, Proceedings Fourth Alvey Vision Conference, Manchester, 1988, 147-151
[19] Ajinkya Chavan, Sreela Sasi, Vision-based Real-Time Speed Tracking, IEEE, 1-3 September, 2005. Faro, Portugal: 16-21.
[20] 高春鸣，兰秋军．应用数字图像增强技术处理医学影像的研究．计算机工程，2001．5
[21] 卢荣胜．视觉准直在线测量技术研究．天津大学精密仪器与光电子工程学院．2000．
[22] 徐奕，周军，周源华．立体视觉匹配技术．计算机工程与应用，2003(15)：1-5．
[23] 边缘和灰度信息结合的SAR图像配准方法研究[D]．北京：中国科学院电子学研究所，2003．
[24] T. Y. Tsai. An efficient and accurate camera calibration technique for 3D machine vision. IEEE Proceedings of Conf. on Computer Vision and Pattern Recognition[C], 1986: 364-374.
[25] T. S. Huang, A. N. Netravali. Motion and structure from feature correspondence: a review, Proceedings of the IEEE, Vol. 82, Vol. 2, 1994: 252-268.
[26] J. Weng, P. Cohen and M. Hemiou. Camera calibration with distortion models and accuracy evaluation. IEEE Trans. Patt. Anal. Machine Intell., 14(10), 1992: 965-980.
[27] 毛剑飞，邹细勇，诸静．改进的平面模板两步法标定摄像机．中国图象图形学报，2004，9(7)：846-852．
[28] Luca Lucchese, Geometric calibration of digital cameras through multi-view rectification. Image and Vision Computing 23 (2005): 517-539.
[29] Joaquim Salvi, Xavier Armangue, Joan Batlle. A comparative review of camera calibrating methods with accuracy evaluation. Pattern Recognition, 2002, 35(1):1617-1635.
[30] 张勇斌．基于单幅编码图像三维场景不标定欧氏重构技术．合肥工业大学博士论文．
[31] 戴丽，顾建刚．最小二乘圆法及其在原木定心中的应用．研究设计，2002(2)．
[32] J. G. Semple and G. T. Kneebone. Algebraic projective geometry[M]. Oxford University Press, New York, 1998: 23-49.
[33] D. Q. Huynh, R. A. Owens, Hartmann, PE. Calibrating a structured light stripe system: a novel approach. International Journal of Computer Vision, Vol.33, No. 1, 1999: 73-86.
[34] O. Faugeras. Stratification of three-dimensional vision: projective, affine, and metric Representations. Journal of the Optical Society of America A, Vol. 12, No. 3, 1994: 465-484.
[35] C. Edberg, A. Ericsson, 3D reconstruction from uncalibrated images[D], Lund Institute of Technology, Lund University, Sweden, December, 2000: 19-79.
[36] Z. Y. Zhang, Determining the Epipolar Geometry and its Uncertainty: A Review. The International Journal of Computer Vision, 27(2): 161-195, Also Research Report No. 2927, INRIA Sophia-Antipolis, March 1998.
[37] M. Pollefeys, L. Van Gool. A stratified approach tometrie self-calibration. Proceedings of the Conference on Computer Vision and Pattern Recognition, Puerto Rico, USA IEEE Computer Society Press, 1997: 407-412.
[38] Ame Henrichsen. 3D reconstruction and camera calibration from2 D images. University of Cape Town, 2000:7-24, 73-88.
[39] R. I. Hartley, In defence of the 8-point algorithm. IEEE Proc. of Fifth International Conference on Computer Vision, 1995:1064-1070.
[40] M. Pollefeys. Self-calibration and metric 3D reconstruction from uncalibrated image sequences. Ph D thesis, ESAT-PSI, K. U. Leuven, 1999.
[41] 施晓红，周佳．精通GUI图形界面编程．北京大学出版社，2003．