本篇論文中，我們提出一套基於攝影學基本理論的曝光校正方法。近年來因為科技的快速發展，手持數位設備例如數位相機也逐漸成為人們手中的口袋產品，在此同時，使用者有著越來越多的機會去使用相機記錄日常生活的點滴，但是這也增加了許多不被使用者滿足的影像，需要透過工具或演算法更進一步的做校正或加強影像品質。
在影像成像的過程中，曝光無非是最重要的一個環節，擁有正確的曝光更往往決定了一張相片的品質好壞。過去許多研究試著在影像品質上做各種程度或目的性的加強，但卻鮮少提及到影像曝光這一個重要的環節。而在日常生活中，曝光不足的情形卻常常發生。像是晚上所拍攝的影像或是背光影像等。
因此，我們基於攝影學上的操作技巧，以及電腦視覺相關的理論來改善影像當中發生曝光不足的情形。我們首先針對影像的畫素值，使用Retinex把它拆解成背景曝光值與反射率兩個構面，然後使用著名的Zone system來標記影像暗部區域中曝光不足的程度。緊接著，我們使用拍攝影像上曝光補償的技巧，根據其對應的曝光程度進行校正。另外一方面，我們也透過在影像反射率的操作，進一步的增加影像中局部區域的對比與細節，使得影像邊緣能就在曝光校正後仍然清楚的保留下來。
在最後的實驗結果中，我們證實本篇提出的方法在跟其他暗部影像增加演算法進行比較時，能夠保有更好的影像品質，包含整體色彩的偏移程度、影像中物體的細節與對比、更自然的色彩。而因為本篇提出的方法在運算複雜度上效率快速，因此我們也實作在晚上所拍攝的視訊畫面上，來證實經過曝光校正後的影片更適合用來觀測、紀錄，或是其他用途。

Exposure is critical in the imaging process, since if the lighting condition is poor, the quality and visibility of images will be severely degraded. In this paper, we propose a fully automatic image enhancement algorithm that can efficiently solve the under-exposed problem. The basic idea behind this method comes from the common technique of photography: the exposure compensation. For poorly exposed images like nighttime images and backlit images, they usually have under exposed regions that directly cause the loss of detail visibility. In order to enhance the badly-exposed image, we first decomposed the image into the lightness and reflection components, and then each of it will be further enhanced separately. We examine the under exposed regions and by adapting the photographic Zone theory to determine the level of exposure at each pixel. Every pixel afterwards is enhanced by the concept of exposure compensation. Besides, the image contrast and edge details are also improved by enhancing the reflection layer. The experiments demonstrate that the proposed algorithm is more effeteness than the other comparatives, which not only include more nature color and visible details, but also induce less unpleased artifacts. In addition, the proposed work can be applied on much more variation cases of badly-exposed images without any manually parameter setting. We also show the efficiency of this work that can be expected for real-time application.

Content
摘要 …………………………………………………………………………………. 4
Abstract …………………………………………………………………...…………. 5
Chapter 1 Introduction ………………………………………………………………. 8
1.1 Research Background …………………………………………...…………… 8
1.2 Motivation and Objective ……………………………………………………10
1.3 Thesis Organization ………………………………………………..………12
Chapter 2 Related Work …………………………………………………….……… 13
2.1 Histogram-based Method ………………………………………….……… 13
2.2 Retinex-based Method …………………………………………………..… 16
2.3 Other Dark Image Enhancement Method ………………………….……… 18
2.4 Zone Theory ………………………………………………………….…… 20
Chapter 3 Proposed Method ………………………………………………...……… 22
3.1 Image Decomposition ………………………………………..…………… 23
3.2 Exposure Enhancement …………………………………………………… 26
3.3 Reflection Enhancement …………………………………………..……… 30
3.4 Balancing Control ………………………………………………………… 35
Chapter 4 Experimental Results ………………………………………….………… 37
4.1 Histogram-based Method ……………………………………….………… 37
4.2 Retinex-based Method ………………………………………….………… 41
4.3 Others existing works ………………………………………………….… 44
4.4 Computational efficiency………………………………………………….45
4.5 Video Enhancement …………………………………………...……...…… 47
Chapter 5 Conclusion …………………………………………………….………… 48
Reference …………………………………………………………………………….49

[1] Rafael C. Gonzalez, and Richard E. Woods, “Digital image processing, 3/E”, Prentice Hall, New Jersey, 2008.
[2] Kim, Yeong-Taeg. "Contrast enhancement using brightness preserving bi-histogram equalization," Consumer Electronics, IEEE Transactions on, vol. 43, Issue 1, pp. 1-8, Feb., 1997.
[3] Wang, Yu, Qian Chen, and Baeomin Zhang, "Image enhancement based on equal area dualistic sub-image histogram equalization method," Consumer Electronics, IEEE Transactions on, vol. 45, Issue 1, pp. 68-75, Feb., 1999.
[4] Chen, Soong-Der, and Abd Rahman Ramli, "Contrast enhancement using recursive mean-separate histogram equalization for scalable brightness preservation," Consumer Electronics, IEEE Transactions on, Vol. 49, Issue 4, pp. 1301-1309, Nov. 2003.
[5] Sim, K. S., C. P. Tso, and Y. Y. Tan, "Recursive sub-image histogram equalization applied to gray scale images," Pattern Recognition Letters, Vol. 28, Issue 10, pp. 1209-1221, July, 2007.
[6] Zuiderveld, Karel , “Contrast limited adaptive histogram equalization,” Graphics gems IV. Academic Press Professional, Inc., 1994.
[7] S. M. Pizer, E. P. Amburn, J. D. Austin, et al.,”Adaptive Histogram Equalization and Its Variations,” Computer Vision, Graphics, and Image Processing, Vol. 39, pp. 355-368, 1987.
[8] Huang, S., F. Cheng, and Y. Chiu, "Efficient Contrast Enhancement Using Adaptive Gamma Correction with Weighting Distribution," Image Processing, IEEE Transaction on, Vol. 22, Issue 3, pp. 1032-1041, March, 2013.
[9] Xuan Dong, Guan Wang, et al., “Fast efficient algorithm for enhancement of low lighting video,” Multimedia and Expo (ICME), 2011 IEEE International Conference on, pp. 1-6, July, 2011.
[10] Land, Edwin H., and John J. McCann, "Lightness and retinex theory," Journal of the Optical society of America, Vol. 61, Issue 1, pp.1-11, 1971.
[11] Rahman, Zia-ur, Daniel J. Jobson, and Glenn A. Woodell, "Multi-scale retinex for color image enhancement," Image processing, International Conference on, Vol. 3, pp. 1003-1006, Sep., 1996.
[12] Peng, Bo, et al., "A new method based on multi-scale retinex for low contrast image enhancement," Fourth International Conference on Digital Image Processing (ICDIP), May, 2012.
[13] Li, Bo, Shuhang Wang, and Yanbing Geng, "Image enhancement based on retinex and lightness decomposition," Image Processing (ICIP), IEEE
50
International Conference on, pp. 3417-3420, Sep., 2011.
[14] He, Kaiming, Jian Sun, and Xiaoou Tang, “Single image haze removal using dark channel prior,” Pattern Analysis and Machine Intelligence, IEEE Transactions on, Vol. 33, Issue 12, pp. 2341-2353, Dec., 2011.
[15] Zhang, Xiangdong, et al., "Enhancement and noise reduction of very low light level images," Pattern Recognition (ICPR), International Conference on, pp. 2034-2037, Nov., 2012.
[16] Fu, Huiyuan, Huadong Ma, and Shixin Wu, "Night Removal by Color Estimation and sparse representation," Pattern Recognition (ICPR), International Conference on, pp. 3656-3659, Nov., 2012.
[17] Chen, Shaohua, and Azeddine Beghdadi, "Natural enhancement of color image," Journal on Image and Video Processing, Jan., 2010.
[18] Raskar, Ramesh, Adrian Ilie, and Jingyi Yu, "Image fusion for context enhancement and video surrealism," ACM SIGGRAPH, 2005.
[19] Yamasaki, Akito, et al., "Denighting: Enhancement of nighttime images for a surveillance camera," Pattern Recognition (ICPR), International Conference on, pp. 1-4, Dec., 2008.
[20] Adams, Ansel.,” The Negative: Exposure and Development,” Ansel Adams Basic Photography Series/Book 2. Boston: New York Graphic Society. ISBN 0-8212-0717-2
[21] C. J. Bartleson and E. J. Breneman,” Brightness perception in complex fields,” Journal of the Optical Society of America, Vol. 57, Issue 7, pp. 953-956, 1967.
[22] Yuan, Lu, and Jian Sun, "Automatic exposure correction of consumer photographs," Computer Vision (ECCV), Vol. 7575, pp. 771-785, 2012.
[23] Ella Parkinson, “The Zone System”, Retrieved July 8, 2013, from the World Wide Web: http://ellaparkinsonphotography.wordpress.com/2012/11/27/the-zone-system/
[24] “Image of zone system”. Retrieved July 8, 2013, from the World Wide Web: http://ellaparkinsonphotography.files.wordpress.com/2012/11/zonesystemerikheyninck.gif
[35] Norman Koren, “A Simplified Zone System”. Retrieved July 8, 2013, from the World Wide Web: http://www.luminous-landscape.com/tutorials/zone_system.shtml