3D立體視覺的廣義來說由兩個階段所組成的。前期階段最主要的目的就是將輸入的影像的，利用我們的立體匹配（Stereo Matching）來求得立體視差值（ Disparity ）再利用簡單的相似三角型公式得到我們的深度圖（Depth Map）。後期階段則是將前面所述產生的深度圖利用depth-based image rendering (DIBR)來產生多個視角的影像。立體匹配是利用大量的運算來找出兩張不同的影像之間所相對應的深度圖。在本論文中我們實作出三種不同的立體匹配方法。pixel-based, window-based, and dynamic programming (DP)-based. Pixel--based 和 window-based 的演算法是屬於區域性的最佳化匹配，而DP則是屬於全域性的最佳化匹配。主要是由三個主要步驟構成：匹配成本的計算（matching cost computation），成本累積（cost aggregation），以及最後是回溯的部分（back-tracing）以硬體實作動態規劃（DP）的立體匹配往往需要大容量的記憶體空間來存取中間計算結果，使得面積無法縮減。在我們的論文裡提出了tile-based 的動態規劃方法，我們將原本的影像切成較小的tile，因此我們在計算差值時只需要較小的記憶體空間。

General 3D stereo vision is composed of two major phases. In the first phase, an image and its corresponding depth map are generated using stereo matching. In the second phase, depth-based image rendering (DIBR) is employed to generate images of different view angles. Stereo matching, a computation-intensive operation, generates the depth maps from two images captured at two different view positions. In this thesis, we present hardware designs of three different stereo matching methods: pixel-based, window-based, and dynamic programming (DP)-based. Pixel--based and window-based methods belong to the local optimization stereo matching methods while DP, one of the global optimization methods, consists of three main processing steps: matching cost computation, cost aggregation, and back-tracing. Hardware implementation of DP-based stereo matching usually requires large memory space to store the intermediate results, leading to large area cost. In this thesis, we propose a tile-based DP method by partition the original image into smaller tiles so that the processing of each tile requires smaller memory size.

第1章 緒論 1
1.1 研究動機 1
1.2 系統架構 1
1.3 論文架構 4
第2章 研究背景與相關研究 5
2.1 研究背景：人類視覺成像 5
2.1.1 人眼與立體成像之關係 5
2.1.2 立體視覺研究背景 8
2.2 相關研究：三維圖像顯示器 10
2.2.1 Lenticular Display 11
2.2.2 Parallax Barrier 11
2.3 DIBR Warping 13
2.3.1 Virtual Image Feature 15
第3章 演算法及設計原理 18
3.1 影像匹配 18
3.1.1 Matching Cost 20
3.1.2 Cost Aggregation 21
3.1.3 Disparity 25
3.2 Local Stereo Matching 26
3.2.1 Pixel-based Stereo Matching 27
3.2.2 Window-based Stereo Matching 28
3.3 Global Stereo Matching 31
3.3.1 Dynamic Programming Algorithm 31
第4章 架構實現與總體最佳化 33
4.1 Local Stereo Matching實作架構 33
4.1.1 Pixel – base Stereo Matching 33
4.1.2 Window – based Stereo Matching 37
4.2 Global Stereo Matching實作架構 39
4.2.1 Dynamic Programming 39
4.3 提出實作方法：Tile – based 43
4.4 效能分析與比較 47
第5章 結論與未來展望 49
5.1 結論 49
5.2 未來展望 49
參考文獻 51

[1] B. Curless and M. Levoy, “A volumetric method for building complex models fromrange images,” ACM Conf. Computer Graphics, Proc. SIGGRAPH ’96, pp. 303-312,Aug. 1996.
[2] M. D. Wheeler, Y. Sato, and K. Ikeuchi, “Consensus surfaces for modeling 3D objects from multiple range images,” in Proc. IEEE ICCV '98, pp. 917-924, Jan.1998.
A. Hilton, A. J. Stoddart, J. Illingworth, and T. Windeatt, “Reliable surface
[3] reconstruction from multiple range images.” in Proc. Eur. Conf. on Computer Vision,pp 117-126, 1996.
[4] S. Seitz and C. Dyer, “Photorealistic scene reconstruction by voxel coloring,” inProc. IEEE Int. Conf. Computer Vision and Pattern Recognition, pp. 1067-1073, Jun.1997.
[5] S. Vedula, P. Rander, H. Saito, and T. Kanade, “Modeling, combining, and rendering dynamic real-world events from image sequences,” in Proc. 4th Conf.
Virtual Systems and Multimedia, vol. 1, pp. 326-332, 1998.
[6] S. J. Gortler, R. Grzeszczuk, R. Szeliski, and M. F. Cohen, “The lumigraph,” ACMConf. Computer Graphics, Proc. SIGGRAPH ’96, pp. 43-54, Aug. 1996.
[7] A. Katayama, K. Tanaka, T. Oshino, and H. Tamura, "A viewpoint dependent stereoscopic display using interpolation of multi-viewpoint images," Proc. SPIE,Stereoscopic Displays and Virtual Reality Systems II, vol. 2409, pp. 11-20, 1995.
[8] S. Laveau and O. Faugeras, “3-D scene representation as a collection of images,” in Proc. International Conference on Pattern Recognition, vol. 1, pp. 689-691, Oct.1994.
[9] M. Levoy and P. Hanrahan, “Light field rendering,” ACM Conf. Computer Graphics,Proc. SIGGRAPH ’96, pp. 31-42, Aug. 1996.[8] S. E. Chen and L. Williams, “View interpolation for image synthesis,” ComputerGraphics and Interactive Techniques, Proc. of SIGGRAPH '93, pp. 279-288, Aug.1993.
[10] L. McMillan and G. Bishop, “Plenoptic modeling: an image-based rendering system,” Computer Graphics, Proc. of SIGGRAPH '95, pp. 39-46, Aug. 1995.
[11] S. Seitz and C. Dyer, “View morphing,” in Proc. of SIGGRAPH '96, pp. 21-30,1996.
[12] C. Vazquez, W. J. Tam and F. Speranza, “Stereoscopic Imaging: Filling Disoccluded Areas in Depth Image-Based Rendering,” in Proc. Of SPIE vol. 6392, 2006.
[13] J. Compen, “3D graphics rendering for multi-view displays: Using programmable shaders on graphics cards,” Koninklijke Philips Electronics N.V. 2005, Technical note TN-2004/00920.
[14] Y.-C. Wu, “Image Synthesis Technology for Multi-view 3D Displays,” Master’s Thesis, Dept. of Electrical Engineering, National Cheng Kung University, 2008.
[15] N. A. Dodgson, “Autostereoscopic 3D Displays,” Computer, vol. 38, iss. 8, pp. 31-36, Aug. 2005.
[16] Daniel S. and Richard S., “A Taxonomy and Evaluation of Dense Two-Frame Stereo Correspondence Algorithms”, International Journal of Computer Vision ,pp 7-42,2001.
[17] B.-G Nam, H. Kim, and H-J Yoo, “A Low-Power Unified Arithmetic Unit For Programmable Handheld 3-D Graphics Systems”, IEEE JOURNAL OF SOLID-STATE CIRCUITS (JSSC), VOL. 42, pp. 1767-1778, NO. 8, AUGUST 2007.
[18] Ruigang Y. and Marc P., “Multi-Resolution Real-Time Stereo on Commodity Graphics Hardware”, Proceedings of the 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’03), pp I-211-217
[19] N. Anantrasirichai, C. Nishan Canagarajah , David W. Redmill and David R. Bull, “Dynamic Programming for Multi-View Disparity/Depth Estimation”, ICASSP 2006 pp II-269-272
[20] S. Park, H. Jeong, “Real-time stereo vision FPGA chip with low-error rate,” in Proc. Int. Conf. Multimedia Ubiquitous Eng., Apr. 2007,pp. 751–756.
[21] Y.-C. Wu, “Image Synthesis Technology for Multi-view 3D Displays,” Master’s Thesis, Dept. of Electrical Engineering, National Cheng Kung University, 2008.
[22] Chul-Hwan Kim*, Ho-Keun Lee, and Yeong-Ho Ha, “Disparity space image based stereo matching using optimal path searching” Proceedings of SPIE–IS&T Electronic Imaging,SPIE Vol. 5022 ,2003.
[23] M. Z. Brown, D. Burschka, and G. D. Hager, “Advances in computational stereo”, IEEE Trans. Pattern Anal. Mach. Intell., vol. 25, no. 8,pp. 993–1008, Aug. 2003.
[24] D. Scharstein and R. Szeliski, “A taxonomy and evaluation of dense two-frame stereo correspondence algorithms,” Int. J. Comput. Vision,vol. 47, pp. 7–42, 2002.
[25] Chun-Jen Tsai and Aggelos K. Katsaggelos, Fellow, IEEE “Dense Disparity Estimation with aDivide-and-Conquer DisparitySpace Image Technique” IEEE TRANSACTIONS ON MULTIMEDIA, VOL. 1, NO. 1, MARCH 1999
[26] Johan Compen, “3D graphics rendering for multiview displays: Using programmable shaders on graphics cards,” Koninklijke Philips Electronics N.V. 2005, Technical note TN-2004/00920, 2005.
[27] Nelson Yen-Chung Chang, Tsung-Hsien Tsai, Bo-Hsiung Hsu, Yi-Chun Chen,and Tian-Sheuan Chang, “ Algorithm and Architecture of Disparity Estimation With Mini-Census Adaptive Support Weight”IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, VOL. 20, NO. 6, JUNE 2010