在本論文中，我們提出了一個從原始單眼影像與相對應深度圖，製作出雙
眼立體影像的方法。我們的方法可以分為二個步驟，在第一個步驟中我們創造出
二個虛擬的影像，我們使用DIBR的演算法來創造出二眼虛擬影像，且我們也解
決了可見性與雜亂空洞問題；在第二個步驟中我們處理了由第一個步驟所產的空
洞問題，我們使用 image inpainting 將周圍區域的紋路與結構同時延伸至空洞區
域中，而空洞區域永遠都屬於背景區域，所以我們使用切割演算法來區分前景與
背景區域，這一系列的演算法組合起來可以產生人眼視覺上合理的雙眼立體影
像。我們選擇數種不同形態的影像來進行實驗，包含了人工背景、自然背景、相
互遮蓋的物體以及棋盤背景。我們的方法在這些影像上都能產生良好的3D效果。

In this thesis, we present a method for creating stereoscopic image from original
monocular image and corresponding depth image. Our method can be divided into
two steps. In first step, we create two virtual views. We apply depth-image-based
rendering algorithm to create two virtual view images and we solved problems of
visibility and scattered hole. In second step, we deal with the newly exposed area
problem generated by rendering process in step 1. We use an image inpainting method
that can simultaneously propagate texture and structure from surrounding area. The
newly exposed area always belongs to background area and so we use the
segmentation algorithm to distinguish foreground area from background area. The
concatenation of all algorithms can generate visually reasonable stereoscopic image
pair for human eyes. We chose several types of image such as artificial background,
natural background, occluded objects and chessboard background. Our method works
well on all of these images and produces good 3D effect by displaying them on 3D
equipment.

ABSTRACT...................................................................................................................................... - 3 -
CHAPTER 1 - INTRODUCTION ................................................................................................... - 4 -
1.1 OUTLINE OF THIS THESIS .........................................................................................................- 4 -
1.2 MOTIVATION AND OBJECTIVES.................................................................................................- 5 -
1.3 SURVEY AND RELATED WORK..................................................................................................- 5 -
CHAPTER 2 - STEREOSCOPIC IMAGE PAIR GENERATION............................................. - 10 -
2.1 DEPTH-IMAGE-BASED RENDERING........................................................................................- 10 -
2.1.1 3D Image Warping ......................................................................................................... - 11 -
2.1.2 Shift-Sensor Algorithm .................................................................................................. - 12 -
2.2 DISOCCLUSSION ......................................................................................................................- 14 -
2.3 VISIBILITY ............................................................................................................................... 15 -
2.4 SCATTERED HOLE FILLING ....................................................................................................- 16 -
CHAPTER 3 - DISOCCLUDED AREA FILLING ALGORITHM............................................ - 18 -
3.1 CORE OF FILLING ALGORITHM..............................................................................................- 18 -
3.2 FILLING ORDER.......................................................................................................................- 20 -
3.3 CONTOUR OF DISOCCLUDED AREA ........................................................................................- 21 -
3.4 GRADIENT INFORMATION .......................................................................................................- 22 -
3.5 COLOR SPACE TRANSFORM....................................................................................................- 23 -
3.6 SEARCHING ALGORITHM........................................................................................................- 24 -
CHAPTER 4 - EXPERIMENTAL RESULTS............................................................................... - 26 -
4.1 SYSTEM ARCHITECTURE.........................................................................................................- 27 -
4.2 SEARCHING RESULTS ..............................................................................................................- 28 -
4.3 FILLING RESULT......................................................................................................................- 29 -
4.3.1 Image with artificial background .................................................................................. - 31 -
4.3.2 Image with natural background .................................................................................... - 34 -
4.3.3 Image with Overlapped Objects ..................................................................................... - 36 -
4.3.4 Image with chessboard background .............................................................................. - 39 -
CHAPTER 5 - CONCLUSION AND FUTURE WORK.............................................................. - 42 -
5.1 CONCLUSION ...........................................................................................................................- 42 -
5.2 FUTURE WORK........................................................................................................................- 43 -

[1] A. Redert, M. Op de Beeck, C. Fehn, W. IJsselsteijn, M. Pollefeys, L. Van Gool,
E. Ofek, I. Sexton, and P. Surman, “ATTEST—advanced three-dimensional
television system techniques” in Proc. 3DPVT’ 02, Padova, Italy, pp. 313–319,
Jun. 2002.
[2] C. Fehn, K. Hopf and Q. Quante. “Key Technologies for an Advanced 3D-TV
System” In Proceedings of SPIE Three-Dimensional TV, Video and Display III,
pp. 66-80, Philadephia, PA, USA, Oct. 2004.
[3] C. Fehn. “A 3D-TV Approach Using Depth-Image-Based Rendering
(DIBR)” In Proceedings of 3rd IASTED Conference on Visualization, Imaging,
and Image Processing , pp. 482-487, Benalmádena, Spain, Sep. 2003.
[4] S. Curti, D. Sirtori, F. Vella, “3D Effect Generation from Monocular View”,
In IEEE Proc. of - 3DPVT’02 International Symposium on 3D Data Processing
Visualization and Transmission, pp. 550 –553, 2002.
[5] 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,
Jun. 2005.
[6] A.Criminisi, P. Pérez, K. Toyama, “Object Removal by Exemplar-Based
Inpainting” In Conf. Computer Vision and Pattern Recog, CVPR'03, Volume 2,
pp. 721-728, Madison, WI, Jun. 2003.
- 45 -
[7] Criminisi, A., Perez, P., Toyama, K., “Region Filling and Object Removal by
Exemplar-Based Image Inpainting“, IP(13), No. 9, pp. 1200-1212, Sep. 2004.
[8] Rafael C. Gonzalez and Richard E. woods, “Digital Image Processing”, Second
Edition.
[9] Citation. Wojciech Matusik, Hanspeter Pfister, “3D TV: A Scalable System for
Real-Time Acquisition, Transmission, and Autostereoscopic Display of
Dynamic Scenes“, ACM Transactions on Graphics. 24(3) 2004
[10] William and Craig pp. 115 – 164 “Seeing 3D from 2D Images”
[11] 王泓銘, 國立成功大學 電機工程學系“影像處理之物體移除演算法及硬體
架構”, 碩士論文2004
[12] 林峻永, 國立交通大學 電機與控制工程學系“應用於立體顯示器之二維轉
立體靜態影像技術”, 碩士論文2005