臺灣博碩士論文加值系統

紅藍立體影像是一個廣泛用來觀看立體影像的方法之一，它可以簡
單地使用紅藍眼鏡就能看到立體效果，優點在於便宜跟方便取得；然
而，它卻不能完整呈現出原本圖片中該有的顏色，也就是有所謂的顏
色失真；再者，它在觀看時也常會造成鬼影的效果，也就是顏色時而
紅時而藍，取決於你用哪隻眼睛觀看。本論文提出了改善觀看紅藍立
體影像的方法，藉由把紅藍立體影像全彩化且藉由較好的觀看裝置來
達到較佳的觀看感覺。本篇論文的想法是利用紅藍立體影像顏色是互
補的特性，也就是在某一眼影像遺失的顏色資訊可以在另一眼影像中
補回，藉由利用這個特性來把左右兩眼影像缺失的顏色通道補回來。
本篇論文把這個問題設定成標籤問題(labeling problem)，藉由標記顏色
資訊來回復左右眼影像。本篇利用了像素比對、顏色關聯性及影像空
間上的平滑度當作是標記顏色的考量，並且利用了最佳化圖形切割方
法(graph cut) 來解決這個問題。此外，也提出了漸進式回復左右眼影
像的方式，利用左右眼缺少顏色資訊的不同及其回復的難易度來做漸
進式的回復，先回復缺少顏色少的影像，再利用其重建的結果來回復
另一張的影像以增強回復的效果。同時，本篇提出的作法與原本未處
理的紅藍影像以及用直覺的顏色傳遞方法(深度圖方法)給使用者觀看並
做比較。結果顯示我們受測者訪談跟回復結果都比其他兩者還要好，
因為我們我有較少的鬼影以及能夠呈現的色彩其真實性較高。

The anaglyph image, primarily used for red-cyan eyeglasses, has the following flaws: it only reproduces partial colors, often brings retinal rivalry and ghosting to viewers. In this work, we propose a method for anaglyph image re-coloring to enhance the color perception and reduce the visual fatigue. When an anaglyph image is given, a corresponding stereo pair can be derived. The goal of this work is to restore lost colors of the stereo pair and thus
convert the anaglyph into full-color 3D contents. Our main idea is to transfer existing information of the images to each other and we formulate this problem as a labeling problem. We integrate terms of pixel matching, color correlation, and spatial smoothness into the proposed cost function. Graph cuts algorithm is also utilized to solve for the local minimum. We also proposed a progressive restoration scheme due to the restoration difficulty of right and left eye image. Finally, we compare our results with the input anaglyph and the results of a disparity-based (na‥ıve color transfer method) color restoration.
Our user study shows that the results from our method improve the viewing experience for anaglyph contents.

1 Introduction 1
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Assumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 Background 5
2.1 Anaglyph Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Computational Anaglyph . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3 Related Work 9
3.1 Stereo Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.2 Color Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.3 Anaglyph Reverting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4 Stereo Color Transfer Model 12
5 Progressive Restoration Scheme 14
5.1 Right Eye Image Restoration . . . . . . . . . . . . . . . . . . . . . . . . 15
5.1.1 Red Color Prior . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
5.2 Left Eye Image Restoration . . . . . . . . . . . . . . . . . . . . . . . . . 18
5.3 Edge-aware Smoothness cost . . . . . . . . . . . . . . . . . . . . . . . . 21
6 Experiments 24
6.1 A Na‥ıve Color Transfer Method . . . . . . . . . . . . . . . . . . . . . . 24
6.2 User Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
6.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6.4 Limitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
7 Conclusion 33
Bibliography 37

[1] Anaglyph methods comparison. http://3dtv.at/Knowhow/
AnaglyphComparison_en.aspx.
[2] Y. Bando, B.-Y. Chen, and T. Nishita. Extracting depth and matte using a colorfiltered
aperture. In ACM SIGGRAPH Asia 2008 papers, SIGGRAPH Asia ’08,
pages 134:1–134:9, New York, NY, USA, 2008. ACM.
[3] T. Basha, Y. Moses, and S. Avidan. Geometrically consistent stereo seam carving.
In Computer Vision (ICCV), 2011 IEEE International Conference on, pages 1816
–1823, nov. 2011.
[4] Y. Boykov and V. Kolmogorov. An experimental comparison of min-cut/max- flow
algorithms for energy minimization in vision. Pattern Analysis and Machine Intelligence,
IEEE Transactions on, 26(9):1124 –1137, sept. 2004.
[5] Y. Boykov, O. Veksler, and R. Zabih. Fast approximate energy minimization via
graph cuts. Pattern Analysis and Machine Intelligence, IEEE Transactions on,
23(11):1222 –1239, nov 2001.
[6] A. J. Chang, H. J. Kim, J. W. Choi, and K. Y. Yu. Ghosting reduction method for
color anaglyphs. volume 6803, page 68031G. SPIE, 2008.
[7] C. Cusano, F. Gasparini, and R. Schettini. Color transfer using semantic image
annotation. volume 8299, page 82990U. SPIE, 2012.

[8] E. Dubois. A projection method to generate anaglyph stereo images. In Acoustics,
Speech, and Signal Processing, 2001. Proceedings. (ICASSP ’01). 2001 IEEE
International Conference on, volume 3, pages 1661 –1664 vol.3, 2001.
[9] A. Hertzmann, C. E. Jacobs, N. Oliver, B. Curless, and D. H. Salesin. Image analogies.
In Proceedings of the 28th annual conference on Computer graphics and interactive
techniques, SIGGRAPH ’01, pages 327–340, New York, NY, USA, 2001.
ACM.
[10] T.-W. Huang and H.-T. Chen. Landmark-based sparse color representations for color
transfer. In Computer Vision, 2009 IEEE 12th International Conference on, pages
199 –204, 29 2009-oct. 2 2009.
[11] I. A. Ideses and L. P. Yaroslavsky. New methods to produce high quality color
anaglyphs for 3-d visualization. In ICIAR (2), pages 273–280, 2004.
[12] H. Kekre and S. Thepade. Color traits transfer to grayscale images. In Emerging
Trends in Engineering and Technology, 2008. ICETET ’08. First International
Conference on, pages 82 –85, july 2008.
[13] V. Kolmogorov and R. Zabin. What energy functions can be minimized via graph
cuts? Pattern Analysis and Machine Intelligence, IEEE Transactions on, 26(2):147
–159, feb. 2004.
[14] A. Levin, D. Lischinski, and Y.Weiss. Colorization using optimization. ACM Trans.
Graph., 23(3):689–694, Aug. 2004.
[15] D. F. McAllister, Y. Zhou, and S. Sullivan. Methods for computing color anaglyphs.
volume 7524, page 75240S. SPIE, 2010.
[16] F. Pitie, A. Kokaram, and R. Dahyot. N- dimensional probability density function
transfer and its application to color transfer. In Computer Vision, 2005. ICCV 2005.
Tenth IEEE International Conference on, volume 2, pages 1434 –1439 Vol. 2, oct.
2005.
[17] E. Reinhard, M. Adhikhmin, B. Gooch, and P. Shirley. Color transfer between images.
Computer Graphics and Applications, IEEE, 21(5):34 –41, sep/oct 2001.
[18] H. Sanftmann and D. Weiskopf. Anaglyph stereo without ghosting. Computer
Graphics Forum, 30(4):1251–1259, 2011.
[19] D. Scharstein and R. Szeliski. A taxonomy and evaluation of dense two-frame stereo
correspondence algorithms. Int. J. Comput. Vision, 47(1-3):7–42, Apr. 2002.
[20] Y.-W. Tai, J. Jia, and C.-K. Tang. Local color transfer via probabilistic segmentation
by expectation- maximization. In Computer Vision and Pattern Recognition, 2005.
CVPR 2005. IEEE Computer Society Conference on, volume 1, pages 747 – 754 vol.
1, june 2005.
[21] J. Watti. Anaglyphs for three-dimensional viewing. http://nzphoto.
tripod.com/sterea/anaglyphs.htm, 2008.
[22] T. Welsh, M. Ashikhmin, and K. Mueller. Transferring color to greyscale images.
ACM Trans. Graph., 21(3):277–280, July 2002.
[23] A. Woods. Understanding crosstalk in stereoscopic displays.
[24] A. J. Woods and C. R. Harris. Comparing levels of crosstalk with red/cyan,
blue/yellow, and green/magenta anaglyph 3d glasses. volume 7524, page 75240Q.
SPIE, 2010.
[25] A. J. Woods and T. Rourke. Ghosting in anaglyphic stereoscopic images. volume
5291, pages 354–365. SPIE, 2004.
[26] L. Yatziv and G. Sapiro. Fast image and video colorization using chrominance
blending. Image Processing, IEEE Transactions on, 15(5):1120 –1129, may 2006.
[27] J.-C. Yoo and T. Han. Fast normalized cross-correlation. Circuits, Systems, and
Signal Processing, 28:819–843, 2009. 10.1007/s00034-009-9130-7.
[28] M. R. U. Zingarelli, L. A. de Andrade, and R. Goularte. Revglyph - a technique for
reverting anaglyph stereoscopic videos. In Proceedings of the 2012 ACM Symposium
on Applied Computing, SAC ’12. ACM, 2012.