( 您好!臺灣時間:2021/05/16 05:08
字體大小: 字級放大   字級縮小   預設字形  


研究生(外文):Meng_I Chiu
論文名稱(外文):Discrete Wavelet Transform Based Image Compression Using Feature Classified Neural Networks
指導教授(外文):Fang-Hsuan ChengAlvin W. Y. Su
外文關鍵詞:Wavelet TransformImage CompressionNeural Network
  • 被引用被引用:3
  • 點閱點閱:283
  • 評分評分:
  • 下載下載:42
  • 收藏至我的研究室書目清單書目收藏:4
本論文將提出以小波轉換為基礎,利用所產生之小波係數加以分析及處理,以形成特殊的小波係數組做為編碼的目標。以期在統壓縮方式下,進一步提高壓縮比,在影像編碼的過程中,系統分為三部分:第一部分為影像空間的轉換,其利用小波轉換 (Discrete Wavelet Transform)中 wavelet function及 scaling function間相互正交特性,將像素間的關係萃取出來產生小波係數。第二部分為對小波係數處理,其利用各頻帶間相互關連的觀念(Subband),擷取小波係數中的影像特徵,並以類神經網路學習各解析層級間的相互關係之權重值,利用此類神經網路權重值與低解析層級的影像,可估測出與原高解析層級之影像相似度極高的重建高解析層級影像(Predicted High Resolution Image),故當有效的利用重建高解析層級影像與原高解析層級影像之相互關係,便可提高影像之壓縮比。第三部分為編碼方式,本論文中採EZW (Embedded Zerotree Wavelet)編碼方式來對應小波轉換的特性,達到最佳壓縮及品質可變(Quality Scalability)的結果,EZW為一較常用於DWT Based的編碼方法。
論文的實驗結果顯示,利用本論文的高頻估測系統所架構的壓縮模式,提供了可比傳統壓縮法較高的壓縮比且較佳的PSNR(Peak Signal To Noise Ratio)效能。
Digital Image Compression is widely applied to the areas such as digital still camera, image database and image communication and so on.
In this thesis, a Discrete-Wavelet Transform based method is used. The difference from the conventional methods is that the wavelet coefficients are processed to generate another set of coefficients in order to increase the compression ratio. The compression system is divided into three parts: The wavelet transformation to generate the scaling coefficients and the wavelet coefficients; A Feature Based Neural Network (FBNN) estimation technique is proposed to estimate the high-resolution wavelet coefficients. Difference Wavelet Coefficients are produced from the original wavelet coefficients and the estimated wavelet coefficients in order to reduce the amount of information to be encoded. Finally, EZW method that is a popular method for encoding the data generated by wavelet decomposition is used to encode all the coefficients for quality and resolution scalability.
The proposed method provides higher compression ration and better PSNR (Peak Signal To Noise Ratio) performance compared to conventional methods in all the computer simulation shown in this thesis.
英文摘要 i
中文摘要 ii
致謝詞 iii
目錄 iv
附圖目錄 vi
附表目錄 x
1 研究背景、動機及目的 1
1.1 研究背景 1
1.2 研究動機 5
1.3 研究目的 6
2 基本研究 12
2.1 離散小波轉換(DWT) 12
2.1.1 一維離散小波轉換與反小波轉換 13
2.1.2 二維離散小波轉換與反小波轉換 20
2.2 類神經網路(Neural Network) 23
2.2.1 類神經網路基本架構 23
2.2.2 誤差返回傳遞(Error-Back Propagation) 26
2.2.3 Simulated Annealing Resilient Back-Propagation SARPROP)類神經網路架構 28
2.3 嵌入式小波轉換壓縮及其解壓縮 31
2.3.1 嵌入式零樹編碼(Embedded Zerotree Coding) 32
2.3.2 EZW編碼之細部原理 34
3 高頻估測系統與影像壓縮系統 36
3.1 高頻估測系統應用於影像壓縮 36
3.2 高頻估測系統 40
3.2.1 類神經網路模型 40 HSARPROP類神經網路架構 40
3.2.2 基本特徵學習向量集合之分類法則 43 影像輪廓邊緣方向性分類 43 高頻變異量分類 46
3.3 高頻估測之於影像放大應用 46
4 實驗結果 50
4.1 高頻估測系統之實驗 50
4.1.1 差值影像資料量與原高解析層級之高頻影像資料量之比較 53
4.1.2 高頻估測之壓縮率比較 56
4.2 高頻估測之於影像放大應用 70
5 結論與未來研究方向 76
參考文獻 77
1. ISO/IEC JTC1/SC29/WG11, “ISO/IEC CD 11172:information technology,” MPEG-1 Committee Draft, Dec. 1991.
2. L. Chiariglione, “Short MPEG-1 Description,” ISO/IEC JTC1/SC29/WG11, Jun. 1996.
3. J. L. Mitchell, W. B. Pennebaker, Chad E.Fogg, and Didier J. LeGall, “MPEG VIDEO COMPRESSION STANDARD,” Chapman&Hall, NY, USA, 1997.
4. ISO/IEC IS 14496. Coding of Moving Pictures and Audio.
5. Masakazu Suzuoki et al., “A Microprocessor with a 128-Bit CPU, Ten Floating-Point MAC’s, Four Floating-Point Dividers, and an MPEG-2 Decoder,” IEEE J. Solid State Circuits, vol. 34, no.11, pp.1608-1618.
6. C.S. Burrus, R.A. Gopinath, and H. Guo. ” Introduction to Wavelet and Wavelet Transforms. Englewood Cliffs,” NJ: Prentice Hall, 1998.
7. Cuy Cote, Berna Erol, Michael Gallant, and Faouzi Kossentini. “H.263+: Video coding at low bit rates,” IEEE Trans. on Circuits Syst. Video Technol., 8(7), November 1998.
8. W. B. Pennebaker and J. L. Mitchell, “JPEG: Still Image Data Compression Standard,” New York: Van Nostran Reinhold, 1993.
9. ISO/IEC JTC1/SC29/WG11, “JPEG-8-R8 Committee Draft,” 1990.
10. ISO/IEC JTC1/SC29/WG1 N1422, “JPEG2000 Verification Model 5.2 (Technical Description),” 1999.
11. B. L. Yeo and B. Liu, “A unified approach to temporal segmentation of motion JPEG and MPEG compressed video,” in Proc. of Int. Conf. on Multimedia Computing and Systems, May 1995, pp. 81-88.
12. ISO/IEC JTC1/SC2/WG11, “Coding of Moving Pictures and Associated Audio For Digital Storage Media at up to 1.5 Mbit/s,” Committee Draft of Standard ISO 11172-2, Nov. 1991.
13. B. B. Chai, J. Vass, and X. Zhuang, “Significance-linked connected component analysis for wavelet image coding,” IEEE Trans. on Image Processing, vol. 8, no. 6, pp. 774-784, 1999.
14. S. D. Serveto, K. Ramchandran and M. T. Orchard, “Wavelet based image coding via morphological prediction of significance,” in Proc. of IEEE International Conference on Image Processing, 1995, pp. 530-533.
15. D. Taubman and A. Zakhor. Multirate 3-D Subband Coding with Motion Compensation. IEEE Transactions on Image Processing, IP-3: 572-588, September 1994.
16. Martin Vetterli and Jelena Kovacevic, Wavelets and Subband Coding, Prentice Hall, 1995.
17. J.M. Shaprio. Embedded Image Coding Using Zerotrees of Wavelet Coefficients. IEEE Transactions on Signal Processing, SP-41: 3445-3462, December 1993.
18. A. Said and W.A. Pearlman. A New Fast and Efficient Coder Based on Set Partitioning in Hierarchical Trees. IEEE Transactions on Circuits and Systems for Video Technologies, June 1996, pages 243-250.
19. D.W. Giffin and J.S. Lim. Muti-Band Exctation Vocoder. IEEE Transcations on Acoustics, Speech, and Signal Processing, 36:1223-1235, August 1988.
20. Martin Vetterli and Jelena Kovacevic, Wavelets and Subband Coding, Prentice Hall, 1995.
21. Setphane Mallat, A Wavelet Tour of Signal Processing, Academic Press, 1998.
22. Raghuveer M. Rao and Ajit S. Bopardikar ”Wavelet Transforms,” Addison Wesley, 1998.
23. Nicholas k. Treadgold and Tamas D. Gedeon, “Simulated Annealing and Weight Decay in Adaptive Learning : The SARPROP Algorithm,” IEEE Trans. Neural Networks, vol. 9,1998, p. 662-667.
24. Judith E. Dayhoff, ”Neural Network Architectures,” International Thomson Computer Press, 1996.
25. Tamura, S.; Tateishi, M., (1997). “Capabilities of a Four-Layered Feedforward Neural Network: Four Layers versus Three,” IEEE Transactions on Neural Networks, 8(2), Mar., pp. 251-255
26. M. Riedmiller and H. Braun, “A direct adaptive method for faster backpropagation learning: The RPROP algorithm,” in Proc. ICNN 93, San Francisco, CA, 1993, pp. 586-591.
27. D. E. Rumelhart, G. E. Hinton, and R. J. Williams, “Learning internal representations bu error propagation,” in Parallel Distributed Processing: Explorations in the Microstructure of Cognition, vol. 1: Foundations, D. E. Rumelhart and J. L. McClelland, Eds. Cambridge, MA: MIT Press, 1986, pp. 318-362.
28. M.F. Moller, “A scaled conjugate gradient algorithm for fast supervised learning,” Neural Networks, vol. 6, 1993, pp. 525-533.
29. M. T. Hugan and M. B. Menhaj, “Training feedforward networks with the Marquardt Qlgorithm,” IEEE Trans. Neural Networks, vol. 5, 1994, p. 989-993,.
30. M.Riedmiller, “RPROP-Description and implementation details,” Univ. Karlsruhe, Germany, Tech.Rep., 1994.
31. Alvin W.Y. Su, and S.F. Liang, “A Generalized Model-Based Analysis/Synthesis Method for Plucked-String Instruments by Using Recurrent Neural Network, ” Presented at the 106th Convention, No.4903, Munich, Germany, 1999.
32. S.F. Liang, and Alvin W.Y. Su, “Recurrent Network Based Physical Model for the Chin and Other Plucked-string Instruments, ” to appear in the November issue in Journal Of Audio Engineering Society (JAES).
33. Jawerth, B. and Sweldens, W., (1994) , “An overview of wavelet based multi- resolution analyses,” SIAM Review, vol. 36, no. 3, pp.377-412.
34. David Taubman, “High Performance Scalable Image Compression with EBCOT,” IEEE TRANSACTIONS ON IMAGE PROCESSING,VOL.9, NO.7, JULY 2000.
35. I. H. Witten, R. M. Neal, and J. G. Cleary, ”Arithmetic Coding for Data Compression,” Communications of ACM, vol. 30, no. 6, 1988, pp. 520-540.
36. Pohsiang Hsu, Beltsville; Kuo Juey Ray Liu, silver Spring, both of Md, “Method And Sustem For Adaptive Video Image Resolution Enhancement,” United States Patent, 464 Apr. 3, 1998, Patent number: 5,991.
第一頁 上一頁 下一頁 最後一頁 top