臺灣博碩士論文加值系統

由於網際網路的發達，使得資訊能不受時間與空間的限制，藉由網路快速的傳播，但是隨著網際網路的急速發展，也產生了許多安全上的問題，例如智慧財產權。在數位化的世界中，複製資料是一件輕而易舉的事，然而未經授權的複製與散佈，對原始創作者來說卻是一大打擊。數位浮水印技術就是目前使用最廣泛，作為多媒體資料保護的一項重要技術。
強健性是數位浮水印技術中一項重要的條件。為了獲得高強健性的系統，我們根據展頻的(spread spectrum)概念，除了最低頻的LL次頻帶以外，將浮水印散佈在各個次頻帶上。首先計算各次頻帶的能量值，依照能量大小分配在該次頻帶所要嵌入的浮水印個數。能量大的次頻帶嵌入較多的浮水印，而能量小的次頻帶，則嵌入較少的浮水印。
在次頻帶中，將浮水印嵌在絕對值較大的小波係數上，並且浮水印的嵌入強度適應於小波係數值的大小，而隨之變化。如此一來，不僅浮水印嵌入強度為最大，也因此獲得極佳的影像品質。浮水印同時以兩種具有互補功能的方式嵌入，使其不論遭遇何種攻擊，均能有其中一種浮水印不被破壞。浮水印因受保護影像遭受攻擊而產生的錯誤部分，再利用通訊上使用的錯誤修正碼加以修正。
本篇論文提出兩種嵌入浮水印的方法，分別是用於不同型式的浮水印。第一種方法可以嵌入Bipolar浮水印，而第二種方法則可以嵌入一張二元影像作為浮水印。一般的二元影像是由黑與白兩種顏色構成，所能包含的資訊較少，因此在這裡利用顫動法，將一張灰階影像，經由顫動法轉換成二元影像，如此一來減少了資料量，卻又能保持灰階大部分的資訊。
為了驗證兩種方法的強健性，在實驗中考慮了許多攻擊法，包括一般的影像處理、JPEG壓縮、StirMark、Jittter以及Photoshop影像處理軟體所提供的影像編輯功能，一共29種攻擊法。實驗結果也顯示，本論文所提出的兩個方法，確實能提供高度的強健性。

The rapid development of Internet makes the multimedia data can be conveyed fast and unlimitedly through the Internet. But it also introduces several security issues such as the intellectual property (IP). Everyone can easily duplicates and distributes the multimedia data in the digital world. But the unauthorized duplication and distribution can violate the rights of the rightful owner. Digital watermarking is a commonly used technique to protect the legal use of the multimedia data.
Robustness is an important requirement in digital watermarking. In order to obtain a digital watermarking system with high robustness, we use the concept of spread spectrum to spread the watermark over subbands, except the lowest frequency band LL. First calculate the energy of each subband and then distribute the watermark among these subbands according to the magnitude of the energy of each subband. If the energy of the subband is large (small), we embed more (less) watermarks.
The watermark is embedded into the most significant wavelet coefficients. The inserting strength of watermark is proportional to the wavelet coefficient and then is adaptive. As a result, we can embed the watermark as strong as possible and obtain better image quality.
The watermark is simultaneously embedded using two different ways that play complementary roles in resisting various kinds of attacks. No matter what kind of attacks it suffers, there is always one watermark can survive. We also use Error-Correcting Code (ECC) to detect and correct the error part of the extracted watermarks.
In this paper, we propose two methods to embed two types of watermark. The first method can embed a bipolar watermark and the second method can embed a binary image. A binary image consists of black and white pixels and can only contain less information. We use the dithering to convert a gray-level image into binary image. Such procedure can reduce the quantity of the data but still can hold most information.
We have considered several attacks including general image processing, JPEG compression, StirMark, Jitter and many image editing accomplished by the PhotoShop tools, totally 29 attacks. Experiment results show that the proposed two methods indeed can provide high robustness performance.

第一章 簡介.................................................1
1.1智慧財產權.............................................1
1.2數位浮水印應滿足的條件.................................2
1.2.1安全性(Security)...................................2
1.2.2強健性(Robustness).................................3
1.2.3透明度(Transparency)或不可視(Imperceptible)........3
1.2.4容量(Capacity).....................................3
1.3數位浮水印基本原理.....................................4
第二章 影像浮水印技術簡介...................................8
2.1 DCT domain頻域法......................................9
2.2.1浮水印結構.........................................9
2.2.2嵌入法設計.........................................9
2.2 Wavelet domain頻域法一...............................12
2.3 Wavelet domain頻域法二...............................15
2.3.1隨機調變法(random modulation).....................15
2.3.2 攻擊對頻域係數的影響.............................16
2.3.3小波係數的選擇....................................17
2.3.4互補調變(complementary modulation)................17
2.3.5浮水印嵌入........................................18
2.3.6浮水印抽取........................................21
2.3.7重新排序程序(relocation)..........................23
第三章 提出的浮水印技術 — 方法一..........................25
3.1浮水印結構............................................25
3.2恰可察覺失真(Just-Noticeable Distortion)..............26
3.3錯誤修正碼(Error-Correcting Code).....................27
3.3.1重複編碼(repetition coding).......................27
3.3.2二元(binary)BCH碼.................................27
3.4小波係數選取..........................................28
3.5浮水印嵌入............................................30
3.6浮水印抽取............................................31
3.7系統評估..............................................32
3.8實驗結果與討論........................................33
3.8.1攻擊法............................................35
3.8.2實驗一：重複碼(t=5)...............................40
3.8.3實驗二：BCH碼.....................................45
3.8.4實驗三：不使用錯誤修正碼..........................47
3.8.5三種實驗比較......................................49
第四章 提出的浮水印技術 — 方法二..........................51
4.1浮水印結構............................................51
4.2小波係數選取..........................................51
4.3浮水印嵌入............................................53
4.4浮水印抽取............................................54
4.5系統評估..............................................55
4.6實驗結果與討論........................................56
4.6.1實驗一：重複碼(t=3)...............................58
4.6.2實驗二：BCH碼.....................................64
4.6.3實驗三：不使用錯誤修正碼..........................67
4.6.4三種實驗比較......................................70
第五章 結論................................................72
5.1重要係數的選取........................................72
5.2適應性嵌入法..........................................73
5.3錯誤修正碼的選擇......................................73
5.4顫動法的應用..........................................73
參考文獻...................................................75

[1]Hartung, F.; Kutter, M. “Multimedia watermarking techniques”, Proceedings of the IEEE, Volume: 87 Issue: 7, July 1999, Page(s): 1079 -1107
[2]I.J.Cox, J.Kilian, F.T.Leighton, and T.Shamoon, “Secure spread spectrum watermarking for multimedia”, IEEE Trans. Image Processing, vol. 6, pp.1673—1687, 1997.
[3]Podilchuk, C.I., Wenjun Zeng, “Image-adaptive watermarking using visual models”, Selected Areas in Communications, IEEE Journal on, Volume: 16 Issue: 4, May 1998 Page(s): 525—539.
[4]R.L.Pickholtz, D.L.Schilling, and L.B.Millstein, “Theory of spread spectrum communications─A tutorial”, IEEE Trans. Commun., vol. COMM-30, pp.855—884, 1982.
[5]Y.S.Kim, O.H.Kwon, R.H. Park, “Wavelet based watermarking method for digital images using the human visual system,” Circuits and Systems, ISCAS ''99. Proceedings of the 1999 IEEE International Symposium on Volume: 4, Page(s): 80 -83 vol.4, 1999.
[6]J.I.Shin and S.H.Park, “MTF and spatial anisotropy based image compression”, IEICE Trans. Fundamentals, Vol.E83-A, No.9, Sep. 2000, Page(s): 1862-1865.
[7]Chun-Shien Lu, Shih-Kun Huang, Chwen-Jye Sze and Hong-Yuan Mark Liao, “Cocktail watermarking for digital image protection”, Multimedia, IEEE Transactions on, Volume: 2 Issue: 4, Dec. 2000, Page(s): 209 —224.
[8]Fabien A. P. Petitcolas, Ross J. Anderson, Markus G. Kuhn., “Attacks on copyright marking systems”, in David Aucsmith (Ed), Information Hiding, Second International Workshop, IH''98, Portland, Oregon, USA, April 15--17, 1998, Proceedings, LNCS 1525, Springer-Verlag, ISBN 3-540-65386-4, pp. 219-239.
[9]Petitcolas, F.A.P.; Anderson, R.J., “Evaluation of copyright marking systems”, Multimedia Computing and Systems, 1999. IEEE International Conference on, Volume: 1, 1999, Page(s): 574 -579 vol.1
[10]A.B.Watson, G.Y.Yang, J.A.Solomon and J.Villasenor, “Visibility of wavelet Quantization noise”, IEEE Trans. Image Processing, vol.6, pp.1164—1175,1997.
[11]A. Hocquenghem, “Codes corecteurs d’erreurs”, Chiffres, 2, pp. 147-156, 1959.
[12]S.Lin, D.J.Costello, Jr., 1983, “Error control coding”, Prentice Hall, pp 142-180.