基於深度繪圖系統是二維圖像轉化為三維立體視覺圖的一種方法。傳統的深度繪圖系統中，輸入端為單視角影像和相對應的深度圖。為了處理單一二維圖像，我們在這篇論文中提出了一種多層的深度繪圖系統。特別是在影像包含毛髮或透明物件時，我們將引用精確摳圖法對其進行處理。本文首先會用精確切割圖的方法將二維影像的物件切出，並產生對應的深度圖。然後再對背景圖進行補洞之後，分別用前景物和背景與相對應的深度圖進行多層的深度繪圖。最後，我們將用合理的層與層的關係把各層進行結合，產生虛擬影像，同時產生三維視覺圖。通過實驗證明，本研究在將單一二維圖像，特別是包含毛髮或透明物件的影像轉化為三維立體圖上有著顯著的效果，並且可大幅減少透明或毛髮物件攜帶的背景資訊，使觀賞者觀看到舒適與高品質的立體視覺圖。

Depth image based rendering (DIBR) is an efficient way of two-dimension (2D) to three-dimension (3D) image conversion. In the traditional DIBR system, the inputs are monoscopic image and its corresponding depth map. In order to deal with a single 2D image, we propose a multi-layer DIBR system. Especially when the image contains hairy or transparent objects, we adopt alpha matting to precisely extract them. First, in the thesis, we use alpha matting method to cut objects from 2D images and generate their corresponding depth maps. Then, after the hole filling for the background, we use the foreground objects and the background to perform the warping process. Finally, we use the reasonable relationship between layers and combine them to create a virtual image such that we could show a high quality three-dimensional stereoscopic image. The experimental results show that the proposed system achieves a success transformation from the 2D images, which are with hairy or transparent objects, to 3D images. This system successfully reduces the background information, which could be carried with hairy or transparent objects and provides the viewer comfortable and high quality stereoscopic images.

[1] J. Sun, J. Jia, C. K. Tang, and H. Y. Shum, “Poisson Matting,” ACM Trans. Graphics, vol. 23, no. 3, pp. 315-321, 2004.
[2] L. Grady, “Random Walks for Image Segmentation,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 28, no. 11, pp. 1768–1783, 2006.
[3] Q. Chen, D. Li, and C. Tang, “Knn Matting,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 35, no. 9, pp. 2175-2188, 2013.
[4] A. Levin, D. Lischinski, and Y. Weiss, “A Closed-Form Solution to Natural Image Matting,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 30, no. 1, pp. 228-242, 2007.
[5] E. S. L. Gastal, and M. M. Oliveira, “Shared Sampling for Real-Time Alpha Matting,” Computer Graphics Forum, vol. 29, no. 2, pp. 575-584, 2010.
[6] J. Wang and M. F. Cohen, “Optimized Color Sampling for Robust Matting,” IEEE Conf. Computer Vision and Pattern Recognition, no. 2, pp. 1-8, 2007.
[7] K. He, C. Rhemann, C. Rother, X. Tang, and J. Sun, “A Global Sampling Method for Alpha Matting,” Proc. IEEE Conf. Computer Vision and Pattern Recognition, pp. 2049-2056, 2011.
[8] A. Torralba and W. T. Freeman, “Properties and Applications of Shape Recipes,” IEEE Conf. Computer Vision and Pattern Recognition, vol. 2, pp. 83-90, 2003.
[9] I. Daribo, and H. Saito, “Bilateral Depth-Discontinuity Filter for Novel View Synthesis,” IEEE International Workshop on Multimedia Signal Processing, pp. 145-149, 2010.
[10] R. Barrett, M. Berry, T. F. Chan, J. Demmel, J. Donato, J. Dongarra, V. Eijkhout, R. Pozo, C. Romine, and H. V. der Vorst, “Templates for the Solution of Linear Systems: Building Blocks for Iterative Methods,” 2nd Edition. SIAM, Philadelphia, PA, 1994.
[11] L. Zhang and W. J. Tam, “Stereoscopic Image Generation Based on Depth Images for 3D TV,” IEEE Trans. on Broadcasting, vol. 51, no. 2, pp. 191-199, 2005.
[12] C. Fehn, “Depth-Image-Based Rendering (DIBR), Compression and Transmission for a New Approach on 3D-TV,” Proc. SPIE Conf. Stereoscopic Displays and Virtual Reality Systems XI, vol. 5291, pp. 93-104, 2004.
[13] Alpha Matting Evaluation Website, http://www.alphamatting.com/datasets.php.
[14] Y. Y. Chuang, A. Agarwala, B. Curless, D. H. Salesin, and R. Szeliski, “Video Matting of Complex Scenes,” ACM Trans. Graphics, vol. 21, no. 3, pp. 243-248, 2002.
[15] Y. Y. Chuang, B. Curless, D. H. Salesin, and R. Szeliski, “A Bayesian Approach to Digital Matting,” IEEE Conf. Computer Vision and Pattern Recognition, vol. 2, pp. 264-271, 2001.
[16] R. Szeliski, “Locally Adapted Hierarchical Basis Preconditioning,” ACM Trans. Graphics, vol. 25, no. 3, pp. 1135-1143, 2006.
[17] P. Lee, and Y. Wu, “Nonlocal Matting,” IEEE Conf. Computer Vision and Pattern Recognition, pp. 2193-2200, 2011.
[18] M. H. Hyun, S. Y. Kim, and Y. S. Ho, “Multi-View Image Matting and Compositing Using Trimap Sharing for Natural 3-D Scene Generation,” IEEE 3DTV Conf. The True Vision – Capture, Transmission and Display of 3D Video, pp. 397-400, 2008.
[19] M. Schmeing and X. Jiang, “Depth Image Based Rendering: A Faithful Approach for the Disocclusion Problem,” 3DTV Conference: The True Vision – Capture, Transmission and Display of 3D Video, pp. 1-4, 2010.
[20] M. Muja and D. G. Lowe, “Fast Approximate Nearest Neighbors with Automatic Algorithm Configuration,” VISAPP International Conf. on Computer Vision Theory and Applications, pp. 171-178, 2001.