臺灣博碩士論文加值系統

在編輯相片時，編輯者通常會想像相片編輯完成後的樣子。這個想
像，經常是受到其他相片的影響，也許是在同一本相簿中的相片或是
之前瀏覽過的相片。但一般使用者時常會遇到想要某種效果，卻不知
道如何達成的問題。基於這樣的觀察，我們發展出了一套易學易用的
相片編輯系統，讓使用者將喜愛且相似主題的相片當做範例，借用範
例照片的顏色來修改照片的顏色。使用者只要用畫筆在兩張照片相對
映的區域做簡單的標記，系統就可以自動的完成編輯的動作。
發展這套系統的目的是希望編輯者能簡單的藉由參考另一張滿意
的照片來修改不滿意的照片。我們設計的使用者介面能讓使用者在兩
張照片相對映的區域畫上標記來轉移想要的顏色。此外，大部份的情
況，使用者只對於一張照片的一部份不滿意。因此，相同的使用流
程，我們的系統也支援局部的修改一張照片，而且我們的演算法能成
功的保存令人滿意部份或是先前修改的區域。
在這篇論文中，我們利用了使用圖分割做最佳化的演算法來找出使
用者希望編輯的區域，也設計了一個新的能量函式，能在計算中同時
考量一個圖素的顏色與位置資訊。另外，我們發展了多重色彩轉移的
方法，用來估計一組轉移函數中各項對一個圖素的影響，並且將這些
影響加總來修改這個圖素。而影像導向的最佳化被用來增加結果中各
區域的一致性，並且消除在邊界上的修改痕跡。
在這篇論文中，我們也展示了我們工具多樣化的應用，並且系統
能產生正確符合使用者期待的結果。更重要的，這套系統非常容易學
習，不需具備要任何攝影或是影像編輯的知識，使用者也不需要改變
照相的習慣。

One of the most common situations in photo editing is that users want some specific effects but don’t know how to achieve them. Nevertheless, they have an image in their mind about how the photo should be, which is derived from a better photo on similar subjects, or from their imagination. With this user motivation in mind, this thesis proposes a method for altering a photo’s color based on the coloring of another exemplary photo. Users simply set the requirement by drawing some pairs of brush strokes in corresponding regions and then the tool will perform the editing automatically.

The goal of our tool is to enable the photo editor to easily alter the dissatisfied photo by referring the other acceptable photo. A suitable user interface have designed for drawing the corresponding regions between a source and one or more targets to transfer the desired color. Besides, in most general cases, users are just dissatisfied with a part of the photo, so our system allow users to edit their photo completely or partially by the same work flow, and our algorithm can preserve the satisfying or ex-editing region successfully.

In this thesis, we find the region the user expect to edit by a graph cut optimization algorithm, and we have designed a new energy function which can consider the color and positional information simultaneously in this process. Additionally, a multiple color transfer method is developed to estimate the different influence of a set of transfer functions on a pixel and accumulate the influences to alter the pixel. Moreover, a image-guide optimization is used to increase the coherence in regions and eliminate the artifacts near edges in the result.

We also display a variety application of our tool in this thesis, and Our system can produce accurate results that match users’ expectations. More importantly, the tool is very easy to learn; no photography or photo editing knowledge is required, and users don’t need to change their habit of taking photos.

摘要i
Abstract iii
1 Introduction 1
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Related work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Our approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 PreviousWork 7
2.1 Color space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Color transfer between images . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Image-guided optimization . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.4 Energy minimization via graph cuts . . . . . . . . . . . . . . . . . . . . 9
3 Example-Based Photo Enhancement 11
3.1 System overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2 Notation definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.3 User interface design . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.4 User requirement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.5 Clamp editing region . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.5.1 Energy function . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.5.2 Color term . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.5.3 Positional term . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.5.4 Smoothness term . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.6 Multiple color transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.7 Global optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4 Result 23
4.1 Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.2 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.3 limitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5 Conclusion and Future Work 35
Bibliography 37

[1] A. Agarwala, M. Dontcheva, M. Agrawala, S. Drucker, A. Colburn, B. Curless, D. Salesin, and M. Cohen. Interactive digital photomontage. International Conference on Computer Graphics and Interactive Techniques, pages 294–302, 2004.
[2] A. Agrawal, R. Raskar, S. Nayar, and Y. Li. Removing photography artifacts using gradient projection and flash-exposure sampling. Proceedings of ACM SIGGRAPH 2005, 24(3):828–835, 2005.
[3] S. Bae, S. Paris, and F. Durand. Two-scale tone management for photographic look. International Conference on Computer Graphics and Interactive Techniques, pages 637–645, 2006.
[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, 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, 2001.
[6] Y. Chang, S. Saito, and M. Nakajima. Example-Based Color Transformation of Image and Video Using Basic Color Categories. Image Processing, IEEE Transactions on, 16(2):329–336, 2007.
[7] Y. Chang, K. Uchikawa, and S. Saito. Example-based color stylization based on categorical perception. Proceedings of the 1st Symposium on Applied perception in graphics and visualization, pages 91–98, 2004.
[8] E. Eisemann and F. Durand. Flash photography enhancement via intrinsic relighting. International Conference on Computer Graphics and Interactive Techniques, pages 673–678, 2004.
[9] J. Jia, J. Sun, C. Tang, and H. Shum. Bayesian correction of image intensity with spatial consideration. ECCV 2004, pages 342–354, 2004.
[10] J. Kopf, M. Cohen, D. Lischinski, and M. Uyttendaele. Joint bilateral upsampling. ACM Transactions on Graphics (Proceedings of SIGGRAPH 2007), 26(3):to appear, 2007.
[11] Y. Li, J. Sun, and H. Shum. Video object cut and paste. Proceedings of ACM SIGGRAPH 2005, 24(3):595–600, 2005.
[12] Y. Li, J. Sun, C. Tang, and H. Shum. Lazy snapping. ACM Transactions on Graphics (TOG), 23(3):303–308, 2004.
[13] D. Lischinski, Z. Farbman, M. Uyttendaele, and R. Szeliski. Interactive local adjustment of tonal values. ACM Transactions on Graphics (TOG), 25(3):646–653, 2006.
[14] G. Petschnigg, R. Szeliski, M. Agrawala, M. Cohen, H. Hoppe, and K. Toyama. Digital photography with flash and no-flash image pairs. ACM Transactions on Graphics (TOG), 23(3):664–672, 2004.
[15] E. Reinhard, M. Adhikhmin, B. Gooch, and P. Shirley. Color transfer between images. Computer Graphics and Applications, IEEE, 21(5):34–41, 2001.
[16] C. Rother, V. Kolmogorov, and A. Blake. ”GrabCut”: interactive foreground extraction using iterated graph cuts. ACM Transactions on Graphics (TOG), 23(3):309– 314, 2004.
[17] D. Ruderman, T. Cronin, and C. Chiao. Statistics of cone responses to natural images: implications for visual coding. Journal of the Optical Society of America A, 15(8):2036–2045, 1998.
[18] Y. Tai, J. Jia, and C. Tang. Local color transfer via probabilistic segmentation by expectation-maximization. Computer Vision and Pattern Recognition, 2005. CVPR 2005. IEEE Computer Society Conference on, 1, 2005.
[19] C.Wang, Q. Yang, M. Chen, X. Tang, and Z. Ye. Progressive cut. Proceedings of the 14th annual ACM international conference on Multimedia, pages 251–260, 2006.