臺灣博碩士論文加值系統

隨著近年來各種多媒體資料於網路上的傳送，資訊隱藏技術亦跟著成長飛速。新一代的影像格式JPEG2000由於有著比JPEG更良好的影像壓縮率與品質，因而有更廣泛之應用。若是在JPEG2000上做資訊隱藏。基於JPEG2000的小波轉換具位元平面編碼的獨特性，使得影像俱有強健性。一定程度內的破壞，還是可以顯示出影像之內容。所以在使用者需求之傳送影像的品質下作資料隱藏是一個值得研究之問題。
JPEG2000小波係數在量化後，JEPG2000使用最佳化區塊壓縮編碼(Embedded Block Coding with Optimized Truncation,EBCOT) 執行過程中會找出每個編碼區塊多個最佳截斷點以將連續的位元資料做分層。依位元失真最佳化截斷點(rate-distortion optimized truncation point)的演算法，會將截斷點與截斷點中間不重要的位元刪除，來達到位元最佳的壓縮率之需求。是以在小波係數量化後做資訊隱藏會造成藏密資料的遺漏。

本論文在JPEG2000小波值域提出的資訊藏密方法是在不採用截斷點最佳化的方式對量化後的係數作資料的隱藏。根據人眼視覺無法詳細辨別影像每個區域的的情形下，對該區域之係數的部份位元，由高位元到低位元依序藏入做藏密的動作，可替代多少位元依視覺效應的計算為主。由於所得到的藏密影像，會有影像斑駁，所以在藏密取代位元前，利用複雜度的計算結果將位於不複雜平面的可取代位元做保留，以達到影像呈現無斑駁的效果。經實驗過後，影像可藏入可觀的資訊且影像的視覺呈現自然且無更動過的痕跡，而在數值分析的影像品質上，也有不錯的表現。

Along with the recent years each kind of multimedia material in network transmission, stegranography was being also rapid with the growth. The next generation for image compression standard JPEG2000 has better compression rate and image quality than JPEG. Steganography with JPEG2000 are more developing. Because of DWT and bit-plane encoded with JPEG2000, the image of JPEG2000 has robustness to bit-errors. The quality of image would not be destroyed seriously by some errors. It is worth to research to steganography with the quality of image under the user demand.
After the quantized DWT wavelet coefficient, JPEG2000 make layer coding in each code block. The layer end point is called truncation point. The EBCOT make find many optimized truncation point in code block with rate-distortion optimized truncation point algorithm to layer coding. That cancel some bit of the middle of two truncation to reach best bit compression rate for decoder. So, the steganography in the quantized DWT wavelet coefficient will make message lost.
In this paper, we proposed steganography in the quantized DWT wavelet coefficient does not use rate-distortion optimized truncation point algorithm. Humans visual do not recognize the detail of every part of image. Steganography in some bit of the wavelet coefficient of the area that humans visual do not recognize clear. The bit replace begin with the least signification bit. The bit of coefficient has to be replaced is base on the human visual calculate. Then show the steganography of the image, will find some patch. To retain the some corresponsive most significant bit prevent the bit replace. The experiment show the image visual nature. The change would not be show in visual.

目 錄
第一章 序論............................................1
第二章 相關背景........................................4
2.1 JPEG2000壓縮格式概觀............................4
2.2 JPEG200之EBCOT格式..............................13
2.3 JPEG2000之現行藏密技術..........................17
2.4 影像品質評估方式................................20
第三章 具視覺遮蔽效應之JPEG2000小波值域之資訊藏密法....21
3.1 視覺遮蔽效應的計算 ..................................24
3.2 複雜度的計算...... ..................................30
3.3 藏密資訊之嵌入..............................36
3.4 藏密之控制資訊..............................40
3.5 隱藏之資料萃取 .............................45
第四章 實驗結果與討論 .................................49
4.1 資料隱藏之實驗結果 ..................................49
第五章 結論與未來展望..................................62

[1]C. Christopoulis, A. Skodres and T. Ebrahimi, “The JPEG2000 Still Image Compression Standard,” IEEE Signal Processing Magazine ,Vol. 8, pp. 36-58, Sept. 2001.

[2]C. Christopoulis, A. Skodres and T. Ebrahimi, ” The JPEG2000 still image coding system: an overview”, Consumer Electronics, IEEE Transactions on Volume 46, Issue 4, Nov. 2000 Page(s):1103 – 1127

[3]H. Noda, J. Spaulding, Mahdad N. Shirazi, M. Niimi, and E. Kawaguchi, “Bit-Plane Decomposition Steganography Combined with JPEG2000 Compression”, F.A.P Prtitcolas IH 2002, LNCS 2578, pp.2578 295-309, 2003.

[4]H. Noda; J. Spaulding; M.N. Shirazi, ; E. Kawaguchi; “Applicationof bit-plane decomposition steganography to JPEG2000 encoded images” IEEE Trans. Volume 9, ISSUE 12, Dec. 2002 Page(s):410 – 413 Digital Object Identifier 10.1109/LSP.2002.806056

[5]H. Zhang; X. Zang, ; J. Liu, “A Secure BPCS Steganography against Statistical Analysis” IEEE CNF. Digital Object Identifier

[6]http://jj2000.epfl.ch/

[7]J. Chen, T. S. Chen and C. Y. Cheng, “A New Scheme of Image Data Hiding Based on EBCOT of JPEG2000 Lossy Compression”, Proceeding of the 2004 IEEE International Conference on Networking, Sensing & Control Taipei, Taiwan, March, pp. 21-23, 2004.

[8]JPEG官方網站http://www.jpeg.org/jpeg2000/index.html

[9]M. Rabbani and D. S. Cruz, “The JPEG 2000 Still-Image Compression Standard” (http://jj2000.epfl.ch/jj_publications/papers/011.pdf)

[10]M. A. Suhail and W. L. Branch, “On the Digital Watermarking in JPEG2000”,Circuits and Systems, 2001.ICECS 2001. The 8th IEEE International Conference on Volume 2,2-5 pp.871-874 vol.2, Sept.2001.

[11]P. C. Su and C. C. Jay Kuo, “Steganography in JPEG2000 Compressed Images”, IEEE Transactions on Consumer Electronics, Vol.49, No.4, November 2003

[12]T. S. Chen, J. Chen and J. G. Chen, “Tamper Detection and Retrieval Technique Based on JPEG2000 with LL Subband”, Networking sensingand Control, 2004 IEEE International Conference on Volume 2, pp.1235 - 1240 Vol.2, 2004

[13]Y. Xiaoyi ; T. Tieniu ; W. Yunhong ; “Reliable detection of BPCS-steganography in natural images”IEEE CNF. Digital Object Identifier 10.1109/ICIG.2004.123. 18-20 Page(s):333 – 336. Dec. 2004

[14]Z. Liang, “Wavelet Domain Steganography for JPEG2000”, IEEE Cons. Volume 1, June 2006 Page(s):40 - 43 Digital Object Identifier10.1109/ICCCAS.2006.284556 10.1109/ICOSP.2006.345711. Volume 2, 16-20 2006.

[15]陳同孝, 張真誠, 黃國峰, ”數位影像處理技術”, pp.1-9, 旗標出版社, 92年出版。

[16]吳炳飛, 胡益強, 瞿忠正, 蘇崇彥, ”JPEG 2000 影像壓縮技術”, 全華圖書出版，92年出版