单帧图像立体显示的算法研究
|
摘要
本文通过立体视觉的理论研究和实验分析，结合计算机图形学、三维重建、图像配准知识提出了一种单帧图像立体增强算法：并编程模拟实现了该算法。文中首先对立体视觉原理深入研究，得出了用于双目投影的数学模型，并分析了模型参数的经验数据。接着对各种常用三维重建算法进行分析比较，着重探讨了适合立体显示技术的线性化SFS算法，并且推导了数学模型公式，提出了相应的改进措施。进一步地，对重建后得到的三维模型，以双目视觉模型为依据，利用基于最大互信息法的图像配准技术生成左右眼图，为立体显示系统提供了数据源。论文最后对全文进行了总结，并对今后的发展方向提出了自己的几点看法。
Based on the discussion of stereoscopic theory and related experiments, this paper gives an stereoscopic display algorithm for single frame image using knowledge of computer graphics、image 3D reconstruction、image registration. Firstly the essential stereoscopic theories such as the formation of stereoscopic perception are introduced in detail, and binocular projection model to calculate stereo pairs are given with key parameters. Secondly, all kinds of traditional 3D image reconstruction algorithms are analyzed and compared. Shape from shading using linear approximation which is most appropriate for stereo display technology was emphasized, its model was created and corresponding improvement was investigated. Furthermore, according to binocular projection model, stereo pairs were created from 3D model which was reconstructed using image registration technology based on maximum mutual information. Finally, the paper draws some conclusions and puts up some proposes about future works.
Based on the discussion of stereoscopic theory and related experiments, this paper gives an stereoscopic display algorithm for single frame image using knowledge of computer graphics、image 3D reconstruction、image registration. Firstly the essential stereoscopic theories such as the formation of stereoscopic perception are introduced in detail, and binocular projection model to calculate stereo pairs are given with key parameters. Secondly, all kinds of traditional 3D image reconstruction algorithms are analyzed and compared. Shape from shading using linear approximation which is most appropriate for stereo display technology was emphasized, its model was created and corresponding improvement was investigated. Furthermore, according to binocular projection model, stereo pairs were created from 3D model which was reconstructed using image registration technology based on maximum mutual information. Finally, the paper draws some conclusions and puts up some proposes about future works.
引文
[1] I. Sexton, P. Surman. Stereoscopic and Autostereoscopic Display Systems [J]. IEEE Signal Processing Magazine, May 1999. 85-99
[2] LYNDON HILL, MEMBER, ADRIAN JACOBS.3-D Liquid Crystal Displays and Their Applications[J].PROCEEDINGS OF THE IEEE, VOL. 94, NO. 3, MARCH 2006: 575-590.
[3] 王元庆．自由立体显示器的应用与现状[J]．现代显示．2003．1：8-11
[4] 邹阳，邓善熙，梁发云．自动立体显示技术的研究发展[J]．合肥工业大学学报．Vol．28 No．8．Aug．2005：869-872．
[5] 吕国强，张宁辉等．体视三维场景浏览器的实现[J]．合肥工业大学学报．vol．9．2005．9：1196-1199．
[6] [日]大石严．显示技术基础[M]．2003：科学出版社．23-54．
[7] 董国耀．透视和体视[M]．1992：北京理工大学出版社．125-350．
[8] Andrew J. Woods. Compatibility of Display Products with Stereoscopic Display Methods[A]. International Display Manufacturing Conference (IDMC) [C], Taiwan, February 2005.
[9] Slinger, C., Cameron, C., Coomber, S. etc. Advances in interactive, holographic 3D displays[J]. IEEE Signal Processing Magazine Yolume 1, 10-14 Nov. 2002 Page(s): 29-30
[10] Michael Halle. Autostereoscopic displays and computer graphics[J]. Computer Graphic. ACM SIGGRAPH, 31(2), 1997. 5: 58-62
[11] NeilA. Dodgson. Autostereoscopic 3D Displays [J].IEEE Computer. August 2005:31-36
[12] Cvan Berket, A R Franklin, J R Mansell. Design and Applications of Multiview 3D-LCD[A]. Proc SID Euro-Display[C]. 96. 1996: 109-112
[13] C van Berkel, J A Clarke. Characterization and Optimization of 3D-LCD module design[A]. Proc SPIE[C]. vol 3012 1997:179-187
[14] 贾正根．多视点成像技术[J]．真空电子技术．2002．2：51-53
[15] 贾正根．立体显示技术新进展[J]．光电子技术．Vol．21 No．4．Dec．2001：267-271
[16] 梁发云，邓善熙，杨永跃．基于裸眼立体显示的视频变换技术研究[J]．应用科学学报．VOL．23，NO．4，2005．7：435-437．
[17] 孙家广．计算机图形学[M]．1998：清华大学出版社．367
[18] 何援军．计算机图形学[M]．2006：机械工业出版社．83-120
[19] 银红霞，杜四春，蔡立军等．计算机图形学[M]．2005：中国水利水电出版社．122-128
[20] 张宇辉，吕国强，胡跃辉，刘文文．立体显示的双目模型算法及实现[J]．计算机应用与工程．2006．
[21] Donald Hearn, M. Pauline Baker．计算机图形学[M]．电子工业出版．2004：246-268．
[22] 梁栋，韦穗，周敏彤．双眼立体感知几何模型的研究[J]．中国图像图形学报．Vol．3．No．8．Aug．1998：679-683
[23] Dudgeon, J. E., Srinivasan, A. Algorithms for graphics texture mapping System Theory[A]. Proceedings. 23rd Southeastern Symposium [C]. 10-12 March 1991. 613-617.
[24] 隋婧、金伟其．双目立体视觉技术的实现及其进展[J]．电子技术应用．2004．10：1-12
[25] 高文、陈熙淼．计算机视觉—算法与系统原理[M]．1999：清华大学出版社：15-17，109-132，187-190
[26] David Marr视觉计算理论．[M]．1988：科学出版社．
[27] R. Basri and W. Jacobs. Lambertian Reflectance and Linear subspaces [J]. IEEE Trans On PAMI, 2003. 25(2): 218-233.
[28] M. Oren and S. K. Nayar, Generalization of the Lambertian model and implications for machine vision, [J]. International Journal of Computer Vision, 1995(14): 227～251.
[29] Ruo zhang, Ping-sing tsai, James Edwin cryer, etc. Shape from shading: A survey[J]. IEEE transactions on pattern analysis and machine intelligence, vol. 21 No. 8, 1999: 690-705
[30] 俞鸿波，赵荣椿，王兵，基于阴影的三维表面重构技术的概述[J]．计算机工程与应用，2004．10：30-33
[31] Horn B K P, Shape from shading: A method for obtaining the shape of a smooth opaque object from one view, [R].1970, M.I.T. Project MAC Rep. TR～79.
[32] Horn B K P, Height and gradient from shading [J]. International Journal of Computer Vision. 1990.5(1): 37～75.
[33] Zheng Q, Chellappa R. Estimation of illuminant direction and shape from shading [J]. IEEETrans. onPAMI. 1991, 13(7):680～702.
[34] Tsai P-S, Shah M. Shape from shading using linear approximation [J].Image and Vision Computing, 199412(8):487～498.
[35] James J. Gibson. The perception of vision surfaces. [J]. The American Journal of Psychology, 1950. 58(3): 367～384.
[36] Y. Ohta. Obtaining surface orientation from texels and perceptive projection [J]Proc. 7th I JCAI: 746-751
[37] J. Aloimonos. Shape from texture. [J]Proc 9th IJCAI: 926-931
[38] R. Bajcsy. Texture Gradient as a depth cue. [J]Computer Graphics and image Process, 5:52-76, 1976.
[39] K. Ikeuchi. Shape from regular patterns [J] Proc Of the5th int joint Conf. On pattern recognition, 1032-1039,1980.
[40] J.R. Kender. Shape from texture: a computational paradigm[J]. Proc. Image Understanding Workshop, 1979: 134-138
[41] A. D. Witkin. Recovering surface shape and orientation from texture [J]AI 17:17-46, 1981.
[42] E. L. Walker and T Kanada. Shape recovering of a solid of resolution from apparent distortions of patterns [J] CMU. Tech. Rep, CMU-CS-80-133,1984.
[43] H. Nakatani. Extraction of vanishing points and its application. [J].Proc. 5″ Int. Conf. On Pattern Recognition, 370-372, 1980.
[44] Canny J. A computation approach to edge detection. [J]IEEE Trans. Pattern Anal. & Math. Intell., 1986, 8(2): 679-698
[45] D.I. Bamea and H.F. Silverman. A Class of Algorithms for Fast Digital Image Registration. [J].IEEE Trans. on computer, 19 72, Vo 1. C-21: 186-192
[46] C.D. MacGillen et al. Image Registration on Variance as a Measure of Overlay Quality. IEEE Traps. Geosci. Electron, Jan1976, vo 1. GE-14, 4419
[47] M. Sved10W, C.D. MacGillen and Paul E. Anuta. Image Registration: Similarity Measure and Preprocessing Method Comparison[J].IEEE Trans on Aerospace and Electronic Systems. 19 92, Vo LAES-14: 141-149
[48] Jan Flussr. An Adaptive Method for Image Registration [J]. Pattern Recognition, 1992, V ol. 25, No. 1: 4 5-54
[49] 周永新，罗述谦．基于形状特征点最大互信息的医学图像配准[J]．2002，14(7)：654—658
[50] 李智，张雅声．基于轮廓特征的图像配准．[J]CCUA图像分会．1997
[51] 郭海涛，刘智，张保明，基于遗传算法的快速影像匹配技术的研究．[J]测绘学院学报．2001(9)
[52] 孔祥勇．基于互信息的多分辨率3D-MR图像配准方法研究．[D]浙江大学．2005
[53] M. Jenkinson and S. M. Smith. Aglobal optimization method for robust affine Registration of brain images.[J].Medical Image Analysis, 2001,5(2):143-156
[54] Frederik Maes, Andre Collignon. Multimodality image registration by maximization of mutual information[J].IEEE Transaction on Medical Imaging, 1997, 16(2):187-198
[55] B. Likar, F. Pernus. A hierarchical approach to elastic registration based on mutual information. [J]. Image and vision computing. 2001,19: 33-44.
[56] Lee C H, Rosenfield A. Improved method of estimating shape from shading using light source coordinate[J]. AI, 1985, 26: 125-143
[2] LYNDON HILL, MEMBER, ADRIAN JACOBS.3-D Liquid Crystal Displays and Their Applications[J].PROCEEDINGS OF THE IEEE, VOL. 94, NO. 3, MARCH 2006: 575-590.
[3] 王元庆．自由立体显示器的应用与现状[J]．现代显示．2003．1：8-11
[4] 邹阳，邓善熙，梁发云．自动立体显示技术的研究发展[J]．合肥工业大学学报．Vol．28 No．8．Aug．2005：869-872．
[5] 吕国强，张宁辉等．体视三维场景浏览器的实现[J]．合肥工业大学学报．vol．9．2005．9：1196-1199．
[6] [日]大石严．显示技术基础[M]．2003：科学出版社．23-54．
[7] 董国耀．透视和体视[M]．1992：北京理工大学出版社．125-350．
[8] Andrew J. Woods. Compatibility of Display Products with Stereoscopic Display Methods[A]. International Display Manufacturing Conference (IDMC) [C], Taiwan, February 2005.
[9] Slinger, C., Cameron, C., Coomber, S. etc. Advances in interactive, holographic 3D displays[J]. IEEE Signal Processing Magazine Yolume 1, 10-14 Nov. 2002 Page(s): 29-30
[10] Michael Halle. Autostereoscopic displays and computer graphics[J]. Computer Graphic. ACM SIGGRAPH, 31(2), 1997. 5: 58-62
[11] NeilA. Dodgson. Autostereoscopic 3D Displays [J].IEEE Computer. August 2005:31-36
[12] Cvan Berket, A R Franklin, J R Mansell. Design and Applications of Multiview 3D-LCD[A]. Proc SID Euro-Display[C]. 96. 1996: 109-112
[13] C van Berkel, J A Clarke. Characterization and Optimization of 3D-LCD module design[A]. Proc SPIE[C]. vol 3012 1997:179-187
[14] 贾正根．多视点成像技术[J]．真空电子技术．2002．2：51-53
[15] 贾正根．立体显示技术新进展[J]．光电子技术．Vol．21 No．4．Dec．2001：267-271
[16] 梁发云，邓善熙，杨永跃．基于裸眼立体显示的视频变换技术研究[J]．应用科学学报．VOL．23，NO．4，2005．7：435-437．
[17] 孙家广．计算机图形学[M]．1998：清华大学出版社．367
[18] 何援军．计算机图形学[M]．2006：机械工业出版社．83-120
[19] 银红霞，杜四春，蔡立军等．计算机图形学[M]．2005：中国水利水电出版社．122-128
[20] 张宇辉，吕国强，胡跃辉，刘文文．立体显示的双目模型算法及实现[J]．计算机应用与工程．2006．
[21] Donald Hearn, M. Pauline Baker．计算机图形学[M]．电子工业出版．2004：246-268．
[22] 梁栋，韦穗，周敏彤．双眼立体感知几何模型的研究[J]．中国图像图形学报．Vol．3．No．8．Aug．1998：679-683
[23] Dudgeon, J. E., Srinivasan, A. Algorithms for graphics texture mapping System Theory[A]. Proceedings. 23rd Southeastern Symposium [C]. 10-12 March 1991. 613-617.
[24] 隋婧、金伟其．双目立体视觉技术的实现及其进展[J]．电子技术应用．2004．10：1-12
[25] 高文、陈熙淼．计算机视觉—算法与系统原理[M]．1999：清华大学出版社：15-17，109-132，187-190
[26] David Marr视觉计算理论．[M]．1988：科学出版社．
[27] R. Basri and W. Jacobs. Lambertian Reflectance and Linear subspaces [J]. IEEE Trans On PAMI, 2003. 25(2): 218-233.
[28] M. Oren and S. K. Nayar, Generalization of the Lambertian model and implications for machine vision, [J]. International Journal of Computer Vision, 1995(14): 227～251.
[29] Ruo zhang, Ping-sing tsai, James Edwin cryer, etc. Shape from shading: A survey[J]. IEEE transactions on pattern analysis and machine intelligence, vol. 21 No. 8, 1999: 690-705
[30] 俞鸿波，赵荣椿，王兵，基于阴影的三维表面重构技术的概述[J]．计算机工程与应用，2004．10：30-33
[31] Horn B K P, Shape from shading: A method for obtaining the shape of a smooth opaque object from one view, [R].1970, M.I.T. Project MAC Rep. TR～79.
[32] Horn B K P, Height and gradient from shading [J]. International Journal of Computer Vision. 1990.5(1): 37～75.
[33] Zheng Q, Chellappa R. Estimation of illuminant direction and shape from shading [J]. IEEETrans. onPAMI. 1991, 13(7):680～702.
[34] Tsai P-S, Shah M. Shape from shading using linear approximation [J].Image and Vision Computing, 199412(8):487～498.
[35] James J. Gibson. The perception of vision surfaces. [J]. The American Journal of Psychology, 1950. 58(3): 367～384.
[36] Y. Ohta. Obtaining surface orientation from texels and perceptive projection [J]Proc. 7th I JCAI: 746-751
[37] J. Aloimonos. Shape from texture. [J]Proc 9th IJCAI: 926-931
[38] R. Bajcsy. Texture Gradient as a depth cue. [J]Computer Graphics and image Process, 5:52-76, 1976.
[39] K. Ikeuchi. Shape from regular patterns [J] Proc Of the5th int joint Conf. On pattern recognition, 1032-1039,1980.
[40] J.R. Kender. Shape from texture: a computational paradigm[J]. Proc. Image Understanding Workshop, 1979: 134-138
[41] A. D. Witkin. Recovering surface shape and orientation from texture [J]AI 17:17-46, 1981.
[42] E. L. Walker and T Kanada. Shape recovering of a solid of resolution from apparent distortions of patterns [J] CMU. Tech. Rep, CMU-CS-80-133,1984.
[43] H. Nakatani. Extraction of vanishing points and its application. [J].Proc. 5″ Int. Conf. On Pattern Recognition, 370-372, 1980.
[44] Canny J. A computation approach to edge detection. [J]IEEE Trans. Pattern Anal. & Math. Intell., 1986, 8(2): 679-698
[45] D.I. Bamea and H.F. Silverman. A Class of Algorithms for Fast Digital Image Registration. [J].IEEE Trans. on computer, 19 72, Vo 1. C-21: 186-192
[46] C.D. MacGillen et al. Image Registration on Variance as a Measure of Overlay Quality. IEEE Traps. Geosci. Electron, Jan1976, vo 1. GE-14, 4419
[47] M. Sved10W, C.D. MacGillen and Paul E. Anuta. Image Registration: Similarity Measure and Preprocessing Method Comparison[J].IEEE Trans on Aerospace and Electronic Systems. 19 92, Vo LAES-14: 141-149
[48] Jan Flussr. An Adaptive Method for Image Registration [J]. Pattern Recognition, 1992, V ol. 25, No. 1: 4 5-54
[49] 周永新，罗述谦．基于形状特征点最大互信息的医学图像配准[J]．2002，14(7)：654—658
[50] 李智，张雅声．基于轮廓特征的图像配准．[J]CCUA图像分会．1997
[51] 郭海涛，刘智，张保明，基于遗传算法的快速影像匹配技术的研究．[J]测绘学院学报．2001(9)
[52] 孔祥勇．基于互信息的多分辨率3D-MR图像配准方法研究．[D]浙江大学．2005
[53] M. Jenkinson and S. M. Smith. Aglobal optimization method for robust affine Registration of brain images.[J].Medical Image Analysis, 2001,5(2):143-156
[54] Frederik Maes, Andre Collignon. Multimodality image registration by maximization of mutual information[J].IEEE Transaction on Medical Imaging, 1997, 16(2):187-198
[55] B. Likar, F. Pernus. A hierarchical approach to elastic registration based on mutual information. [J]. Image and vision computing. 2001,19: 33-44.
[56] Lee C H, Rosenfield A. Improved method of estimating shape from shading using light source coordinate[J]. AI, 1985, 26: 125-143