本論文提出同質紋理之快速視差傳遞立體匹配演算法，此立體匹配演算法分為三個步驟：首先，利用可適性支持權重演算法計算各像素之視差，根據像素間顏色相似度及幾何距離調整權重，以提升計算準確度，並降低計算複雜度，使演算法適合硬體實現。接著，利用視差傳遞演算法根據區塊之同質紋理平滑程度及區塊邊界之視差，判斷是否傳遞可靠視差值至整個區塊，以達到加速之效果，使用此技術約可降低82%的計算時間。最後，利用後處理演算法以左右一致性檢查方式檢測出視差圖中的錯誤點，並分類成不匹配點與遮蔽區域，經由不同後處理法改善視差圖中之錯誤點，以提升視差圖品質。本立體匹配演算法合成後之結果與原視差相當接近，硬體架構約需108 k個邏輯閘，最高之操作頻率可達到20 MHz，支援每秒30張畫面而解析度為384 × 288且視差搜尋範圍為16像素點之圖片。

A stereo matching algorithm with fast disparity propagation under homogeneous texture detection is proposed in this thesis. The algorithm consists of three major components. First, a modified low complexity adaptive support weight stereo matching algorithm is adopted to compute the disparity. It uses fixed-sized support window with adaptive weights based on the color similarity and geometry proximity. Secondly, the disparity propagation algorithm based on homogeneous texture detection is used to reduce the computation. Pixels of block are determined to reduce the disparity hypotheses or propagate the disparity. The proposed algorithm with disparity propagation technique reduces about 82% of the execution time compared to the original algorithm. Final, the occlusion or mismatch errors are determined by left-right consistency check. Different adjusted post-processing algorithms are used to solve these incorrect errors. Synthesis result shows the implemented gate counts is 108k and the maximum operation frequency is 20MHz. It can support 30 fps for image size of 384 × 288 with 16 disparity range levels.

[1] C. Fehn, K. Hopf, and B. Quante, “Key technologies for an advanced 3D TV system,” in Proc. SPIE Three-Dimensional TV, Video, and Display Ⅲ, Oct. 2004, pp. 66-80.
[2] C. Fehn, R. D. L. Barre, and S. Pastoor, “Interactive 3-DTV: concepts and key technologies,” Proc. IEEE, vol. 94, pp. 524-538, Mar. 2006.
[3] L. H. Wang, X. J. Huang, M. Xi, D. X. Li, and M. Zhang, “An asymmetric edge adaptive filter for depth generation and hole filling in 3DTV,” IEEE Trans. Broadcast., vol. 56, pp. 425-431, Sept. 2010.
[4] I. Daribo and H. Saito, “A novel inpainting-based layered depth video for 3DTV,” IEEE Trans. Broadcast., vol. 57, pp.533-541, June 2011.
[5] Y. Kuk-Jin and K. In So, “Adaptive support-weight approach for correspondence search,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 28, pp. 650-656, Apr. 2006.
[6] T. Kanade and M. Okutomi, “A stereo matching algorithm with an adaptive window: theory and experiment,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 16, pp. 920-932, Sept. 1994.
[7] Y. Boykov, O. Veksler, and R. Zabih, “A variable window approach to early vision,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 20, pp. 1283-1294, Dec. 1998.
[8] O. Veksler, “Stereo correspondence with compact windows via minimum ratio cycle,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 24, pp. 1654-1660, Dec. 2002.
[9] Z. Ke, L. Jiangbo, and G. Lafruit, “Cross-based local stereo matching using orthogonal integral images,” IEEE Trans. Circuits Syst. Video Technol., vol. 19, pp. 1073-1079, July 2009.
[10] O. Veksler, “Fast variable window for stereo correspondence using integral images,” in Proc. IEEE Comput. Vision and Pattern Recognition, June 2003, pp. 556-561.
[11] Y. Zhang and C. Kambhamettu, “Stereo matching with segmentation-based cooperation,” in Proc. Eur. Conf. Comput Vision, Copenhagen, Denmark, May 2002, pp. 556-571.
[12] H. Lim and H. Park, “A dense disparity estimation method using color segmentation and energy minimization,” in Proc. IEEE Image Process., Oct. 2006, pp. 1033-1036.
[13] M. Gong and Y. Yee-Hong, “Fast stereo matching using reliability-based dynamic programming and consistency constraints,” in Proc. IEEE Comput. Vision, Oct. 2003, pp. 610-617.
[14] K. Jae Chul, L. Kyoung Mu, C. Byoung Tae, and L. Sang Uk, “A dense stereo matching using two-pass dynamic programming with generalized ground control points,” in Proc. IEEE CVPR, June 2005, pp. 1075-1082.
[15] W. Yichen and Q. Long, “Region-based progressive stereo matching,” in Proc. IEEE CVPR, June 2004, pp. I-106 - I-113.
[16] S. Xun, M. Xing, J. Shaohui, Z. Mingcai, and W. Haitao, “Stereo matching with reliable disparity propagation,” in Proc. IEEE 3DIMPVT, May 2011, pp. 132-139.
[17] D. Scharstein, R. Szeliski, and R. Zabih, “A taxonomy and evaluation of dense two-frame stereo correspondence algorithms,” in Proc. IEEE SMBV, Dec. 2001, pp. 131-140.
[18] D. Scharstein and R. Szeliski, “High-accuracy stereo depth maps using structured light,” in Proc. IEEE CVPR, June 2003, pp. I-195-I-202.
[19] D. Scharstein and C. Pal, “Learning conditional random fields for stereo,” in Proc. IEEE CVPR, June 2007, pp. 1-8.
[20] H. Hirschmuller and D. Scharstein, “Evaluation of cost functions for stereo matching,” in Proc. IEEE CVPR, June 2007, pp. 1-8.
[21] Middlebury Stereo Vision Page [Online]. Availale: http://vision.middlebury.edu/stereo
[22] A. F. Bobick and S. S. Intille, “Large occlusion stereo,” Int. J. of Comput. Vision, vol. 33, pp. 181-200, Sept. 1999.
[23] Y. Boykov, O. Veksler, and R. Zabih, “Fast approximate energy minimization via graph cuts,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 23, pp. 1222-1239, Aug. 2001.
[24] P. F. Felzenszwalb and D. R. Huttenlocher, “Efficient belief propagation for early vision,” in Proc. IEEE CVPR, June 2004, pp. I-261-I-268.
[25] T. Kanade, H. Kano, S. Kimura, A. Yoshida, and K. Oda, “Development of a video-rate stereo machine,” in Proc. IEEE Intell. Robots and Syst., Aug. 1995, pp. 95-100.
[26] C. Georgoulas and I. Andreadis, “A real-time occlusion aware hardware structure for disparity map computation,” In Proc. Image Anal. and Process, Vietri sul Mare, Italy, June 2009, pp. 721-730.
[27] K. Ambrosch, M. Humenberger, W. Kubinger, and A. Steininger, “A SAD-based stereo matching using FPGAs,” In Proc. Embedded Comput. Vision, Kyoto, Japan, 2009, pp. 121-138.
[28] A. Darabiha, J. MacLean, and J. Rose, “Reconfigurable hardware implementation of a phasecorrelation stereo algorithm,” Mach. Vision Appl., vol. 17, pp. 116–132, Apr. 2006.
[29] S. Jin, J. Cho, X. D. Pham, K. M. Lee, S-K. Park, M. Kim, and J. W. Jeon, “FPGA design and implementation of a real-time stereo vision system,” IEEE Trans. Circuits Syst. Video Technol., vol. 20, pp. 15-26, Jan. 2010.
[30] K. Ambrosch and W. Kubinger, “Accurate hardware-based stereo vision,” Comput. Vision and Image Understanding, vol. 115, pp. 1303-1316, Feb. 2011.