由於網路的盛行，資料呈現的形式大多已數位化，大量的多媒體資訊例如影像、聲音…等，透過網路方便、快速的傳輸，使人們得以很快的接收到所需的訊息；如何增快傳輸速度？一般人直觀的想法不外乎加大頻寬或是減低資料量，在加大頻寬方面為滿足使用者的需求必須不斷地擴充頻寬，這對於成本而言並沒有效率，所以減低資料量是較可行的方法，目前已有很多的壓縮方法被提出來。在本篇論文中，我們提出兩個植基於向量量化編碼技術的影像壓縮方法。
在第一個方法中，我們改進向量量化編碼法在編碼時所耗費的時間，每個區塊利用額外一個位元記錄該區塊的像素值是由編碼字或是區塊平均值取代，一般而言每張影像平均有80~90%的區塊不需要透過向量量化編碼來搜尋最短距離的編碼字，因此編碼的時間便減少了，而且不論就視覺上或是數據上，所還原的影像品質與向量量化編碼法相比較皆不遜色，有的影像品質甚至比單純使用向量量化編碼法來得還要好。
在第二個方法中，我們首先將影像做向量量化編碼後，取得索引值，然後再把該索引表做第二次的壓縮，為了達到還原後的影像品質與向量量化編碼法相同，第二次壓縮採用無失真的壓縮方法，利用先前索引值出現次數統計來預測目前要編碼的索引值，加上有限狀態向量量化編碼法中狀態編碼簿的概念配合無失真的編碼技術，根據我們的實驗結果，壓縮率比以前所提出的方法來得好，我們更進一步討論狀態編碼簿大小，利用門檻值來決定使用狀態編碼簿的大小，其大小的範圍為：21~23，經由實驗可知大部分狀態編碼簿大小變動比固定大小的狀態編碼簿所產生的位元比低。

Internet becomes a popular transmission channel recently. Most data such as text, image, audio and video can be represented in digital form and transmit on the network. As we know the bandwidth is restricted, too many multimedia data transmit over Internet result in congestion or delay. The performance of network is degraded. In order to increase the compression rate but preserve more image detail, many image compression methods are proposed in the last decades. Since the human visual system is low sensitivity, we can reduce the image size but still keeping the image quality in allowable range. In this thesis, two methods based on Vector Quantization (VQ) for lossy image compression are proposed.
In the first scheme, we add an extra bit per block to indicate whether the block needs to search the nearest codeword or not. Each block is compressed according to the variance of it. If the variance is less than the predefined threshold, we represented the pixel values in the block with the mean of the block. Otherwise, we choose a nearest codeword to represent the block. We have less calculating time and better image quality than VQ after reconstructed. The experimental results reveal the compression rate of our method is almost as good as VQ.
In the second scheme, we propose a post-processing method for VQ compressed image. We apply lossless encoding technique, predictor and the concept of state codebook for lossless image compression. In our experimental results, the bit rate is lower than previous methods. We discuss the size of state codebook further. For the variable size approach, we set two thresholds to decide the size of state codebook. According to the experimental results we can find that the compression ratio of variable size is better than fixed size.

中文摘要　　……………………………………………………… i
英文摘要　　……………………………………………………… ii
誌謝　　…………………………………………………………… iii
目錄　　…………………………………………………………… iv
圖次　　…………………………………………………………… vi
表次　　…………………………………………………………… vii
第一章　緒論　　………………………………………………… 1
　第一節　研究背景與動機　…………………………………… 1
　第二節　研究目的　　………………………………………… 1
　第三節　文章架構　　………………………………………… 2
第二章　文獻探討　　…………………………………………… 4
　第一節　失真壓縮方法介紹　　……………………………… 5
　第二節　無失真壓縮方法介　　……………………………… 13
　第三節　影像評估方法　　…………………………………… 15
第三章　應用區塊平均值及向量量化編碼法於失真影像壓縮　 17
　第一節　相關研究　　………………………………………… 17
　第二節　提出的方法　　……………………………………… 18
　第三節　實驗結果　　………………………………………… 21
第四章　植基於向量量化編碼技術的二次編碼法　　………… 35
　第一節　相關研究　　………………………………………… 35
　第二節　提出方法　　………………………………………… 39
　第三節　實驗結果　　………………………………………… 45
第五章　結論及未來研究方向　　……………………………… 55
　第一節　結論　　……………………………………………… 55
　第二節　未來研究方向　　…………………………………… 56
參考文獻　　……………………………………………………… 57

[1]張真誠、黃國峰、陳同孝(2000)，電子影像技術，台北市，松崗電腦圖書資料股份有限公司。
[2]陳同孝、張真誠、黃國峰(2001)，數位影像處理技術，台北市，松崗電腦圖書資料股份有限公司。
[3]交通大學自動化資訊處理實驗室網頁
http://debut.cis.nctu.edu.tw/pages/compress.htm
[4]宋建行論文
http://www.ee.yuntech.edu.tw/~dsplab/dsplab/paper/宋建行論文.htm
[5]林達德教授的網頁─教學講義http://nova.bime.ntu.edu.tw/~ttlin/Course01/lecture_notes/c1lecture_note09.htm
[6]M. Accame and F. Granelli (1999), Hierarchical progressive image coding controlled by a region based approach, IEEE Transactions on Consummer Electronics, Vol.146, No.3, pp. 13-20.
[7]J. Askelof, M.L. Carlander, C. Christopoulos (2002), Region of interest coding in JPEG 2000, Signal Processing: Image Communication, pp.105-111.
[8]B.B. Chai, J. Vass and X. Zhuang (1999), Significance-linked connected component analysis for wavelet image coding, IEEE Transactions on Image Processing, Vol.8, No.6, pp.774-784.
[9]C.C. Chang and T.S. Chen (1997), A new tree-structured vector quantization with closest-coupled multipath searching method, Optical Engineering, Vol.36, No.6, pp.1713-1720.
[10]C.C. Chang, J.S. Chou and T.S. Chen (2000), A fast prediction of Euclidean distance using a look-up table, IEEE Transactions on Circuits and Systems for Video Technology, Vol.10, No.4, pp.594-599.
[11]C.C. Chang, C.C. Jau, and T.S. Chen (1999), K-ary TSVQ: a new improvement of tree structured vector quantization, Proceedings of IASTED International Conference Modeling and Simulation (MS’99), Philadephia, USA, pp.253-239.
[12]C.C. Chang, D.C. Lin, and T.S. Chen (1997), An improved VQ codebook search algorithm using principal component analysis, Journal of Visual Communication and Image Represent, Vol.8, No.1, pp.27-37.
[13]C.C. Chang, F.J. Shiue, and T.S. Chen (1999), Tree structured vector quantization with dynamic path search, Proceedings of International Workshop on Multimedia Network Systems (MMNS), Aizu, Japan, pp.536-541.
[14]C.C. Chang, C.H. Sung and T.S. Chen (1997), A locally adaptive scheme for image index compression, Proceedings of the 1997 Conference on Computer Vision, Graphics, and Image Processing, Taichung, Taiwan, pp.93-99.
[15]R.F. Chang, W.T. Chen and J.S. Wang (1992), Image sequence coding adaptive tree structured vector quantization with multipath searching, IEE Proceeding I, Communications, Speech and Vision, Vol.1, pp.9-14.
[16]K.L. Chung and S.Y. Tseng (2001), New progressive image transmission based on quadtree and shading approach with resolution control, Pattern Recognition Letters, pp.1545-1555.
[17]A. Cziho, B. Solaiman, I. Lovanyi, G. Cazuguel, and C. Roux (2000), An optimization of finite-state vector quantization for image compression, Signal Processing: Image Communication, pp.545-558.
[18]G. Deng (2002), Transform domain LMS-based adaptive prediction for lossless image coding, Signal Processing: Image Communication, pp.219-229.
[19]K. Egiazarian (2001), Lossless image compression by LMS adaptive filter banks, Signal Processing, pp.447-450.
[20]R.M. Gray (1984), Vector quantization, IEEE ASSP Magazine, pp.4-29.
[21]H.J. Grosse, M.R. Varley, T.J. Terrell, and Y.K. Chan (2000), Improved coding of transform coefficients in JPEG-like image compression schemes, Pattern Recognition Letters, pp.1061-1069.
[22]C.H. Hsieh and J.C. Tsai (1996), Lossless compression of VQ index with search-order coding, IEEE Transactions on Image Processing, Vol.5, No.1, pp.1579-1582.
[23]C.H. Hsieh, K.C. Chuang and J.S. Shue (1993), Image compression using finite-state vector quantization with derailment compensation, IEEE Transactions on Circuits and Systems for Video Technology, Vol. 3, No. 5, pp.341-349.
[24]J. Huguet and L. Torres (1990), Fast codebook search algorithm in vector quantization, Visual Communication and Image Processing, pp.1750-1755.
[25]A. Jarvi, J. Lehtien and O. Nevalainen (1999), Variable quality image compression system based on SPIHT, Signal Processing: Image Communication, Vol.14, No.9, pp.683-696.
[26]R.Y. Li, J. Kim, and N. Al-Shamakhi (2002), Image compression using transformed vector quantization, Image and Vision Computing, pp.37-45.
[27]Y.C. Liaw, W. Lo and Z.C. Lai (2002), Image restoration of compressed image using classified vector quantization, Pattern Recognition, pp.329-340.
[28]Y. Linde, A. Buzo and R.M. Gray (1980), An algorithm for vector quantization design, IEEE Transactions on Communications, Vol.28, pp.84-95.
[29]Z.J. Liu (1998), Detection and correction of transmission error in JPEG images, IEEE Transactions on Circuits and Systems for Video Technology, Vol.8, No.2, pp.221-231.
[30]M.A. Losada, G. Tohumoglu, D. Fraile and A. Artes (2000), Multi-iteration wavelete zero-tree coding for image compression, Signal Processing, pp.1281-1287.
[31]A. Makur and S.S. Selvi (2001), Variable dimension vector quantization based image watermarking, Signal Processing, pp.889-893.
[32]A. Munteanu, J. Cornelis, G.V.D. Auwera and P. Cristea (1999), Wevelet image compression — the quadtree coding approach, IEEE Transactions on Technology in Biomedicine, Vol.3, No.3, pp.176-185.
[33]S.A. Rizvi and N.M. Nasrabadi (1997), Finite-state residual vector quantization using a tree-structured competitive neural network, IEEE Transactions on Circuits and Systems for Video Technology, Vol.7, No.2, pp.377-390.
[34]S. Servetto, K. Ramchandran, and M.T. Orchard (1995), Wavelet based image coding via morphological prediction of significance, Proceedings of IEEE International Conference on Image Processing, pp. 530-533.
[35]J.M. Shapiro (1993), Embedded image coding using zerotrees of wavelet coefficients, IEEE Transactions on Signal Processing, Vol.41, No.12, pp.3445-3462.
[36]A. Said and W.A. Pearlman (1996), A new, fast, and efficient image codec based on set partitioning in hierarchical trees, IEEE Transactions on Circuits and System for Video Technology, Vol.6, No.3, pp.243-250.
[37]A.N. Skodras, C.A. Christopoulos, T. Ebrahimi (2001), JPEG2000: the upcoming still image compression standard, Pattern Recognition Letters, pp.1337-1345.
[38]H.J. Wang and C.C.J. Huo (1997), A multi-threshold wavelet coder (MTWC) for high fidelity image compression, IEEE Proceedings International Conference on Image Processing, Vol.1, pp.652-655.
[39]B.F. Wu and C.Y. Su (2000), Low computational complexity enhanced zerotree coding for wavelet-based image compression, Signal Processing: Image Communication, pp.401-411.
[40]M. Xu and A. Kuh (1996), Image coding using feature map finite-state vector quantization, IEEE Signal Processing Letter, pp.215-217.
[41]S.B. Yang and L.Y. Tseng (2001), Smooth side-match classified vector quantizer with variable block size, IEEE Transactions on Image Processing, Vol.10, No.5, pp.677-685.