臺灣博碩士論文加值系統

影像壓縮定位在如何降低數位影像資料量的問題上，基本的處理方式就是移除數位影像中累贅的資料。在影像壓縮的研究中，有許多種技術可以使用，像是向量量化，小波轉換、碎形影像壓縮及有名之壓縮標準JPEG、GIF…等等。。
支撐向量機是一種使用高維度特徵空間中的學習系統，它的基本原則是Vapnik在1992年提出的。支撐向量機的演算法有兩種主要的方法 --- 一種是支撐向量分類，另一種是支撐向量迴歸。
在這篇論文裡，我們嚐試使用不同於傳統的方法，即支撐向量機來壓縮灰階影像。我們使用來壓縮影像的主要方法是支撐向量迴歸演算法，利用區塊的支撐向量數量來達到壓縮的目的並且實作驗證。
本文中將支撐向量迴歸使用在影像中的取樣區塊並作分析，以瞭解支撐向量迴歸對取樣區塊的影響。我們提出了二個方法，第一個方法是基本的支撐向量迴歸編碼法，我們將支撐向量迴歸用於影像壓縮。第二個方法是適應性支撐向量迴歸編碼法，在這個方法中，我們改善了取樣區塊的PSNR值及編碼後影像中的區塊效應。

Image compression addresses the problem of how to reduce the amount of data required to represent the image and the basic process is the removal of redundant data in the image. In the researches of image compression, there are many kinds of techniques such as vector quantization (VQ), Wavelet transform, fractal theory, well-known standard of compression --- JPEG, GIF and so on.
Support vector machine (SVM) is a learning system, which was first introduced by Vapnik in 1992. There are two main categories for support vector machines: support vector classification (SVC) and support vector regression (SVR). SVM can generalize complicated gray level structures with only very few support vectors and thus provides a new mechanism for image compression.
In this thesis, the algorithms of SVM are utilized to perform the compression of gray images. The method used is SVR which reduces the number of support vectors of image blocks in order to improve the compression ratio. Meanwhile, the parameters for SVR are selected according to the properties of the block in the frequency domain so as to improve the qualities of the retrieved images.

摘 要 I
Abstract II
誌 謝 III
LIST OF CONTENTS IV
LIST OF TABLES V
LIST OF FIGURES VI
Chapter 1 Introduction 1
Chapter 2 Support vector machine 3
2.1 The learning system 3
2.2 The feature space and kernel functions 4
2.3 Support vector classification (SVC) 5
2.3.1 The primal form 5
2.3.2 The Karush-Kuhn-Tucker (KKT) conditions 6
2.3.3 The dual form 7
2.3.4 The kernel form 9
2.3.5 The 2-norm soft margin classifier 9
2.3.6 The 1-norm soft margin classifier 11
2.4 Support vector regression (SVR) 11
2.5 Predictive function of a non-linear SVR 13
Chapter 3 SVR-based image block coding using MATLAB 15
3.1 Discrete cosine transform (DCT) 15
3.2 Experiment results 16
3.2.1 Sampling blocks and methods 16
3.2.2 Block compression using SVR 19
Chapter 4 The baseline SVR encoding method 24
4.1 The environment 24
4.2 Implementation 24
4.3 Experiment results 27
Chapter 5 The adaptive SVR encoding method 32
5.1 The problems of the baseline SVR encoding method 32
5.2 SAAC and parameters of Gaussian RBF 34
5.3 Implementation 35
5.4 Experiment results 38
5.5 Conclusion 38
Reference 43

[1] Nello Cristianini and John Shawe-Taylor, “An Introduction to Support Vector Machines and other kernel-based learning methods”, Cambridge university, 2002
[2] Pierre M.L. Drezet and Robert F. Harrison, “A new method for sparsity control in support vector classification and regression”, the university of Sheffield, 1999
[3] Chen-Chia Chuang, Jin-Tsong Jeng and Chih-Ching Hsiao, “Support vector regression for image filter and image compression”, 2002
[4] Steve R. Gunn, “Support Vector Machines for Classification and Regression”, Faculty of Engineering and Applied Science and Department of Electronics and Computer Science, 1998
[5] Marti A. Hearst, “Support Vector Machine”, IEEE Intelligent Systems, University of California, Berkeley, July/August 1998
[6] Rafael C. Gonzalez and Richard E. Woods, “Digital Image Processing”, Addison Wesley, 1993
[7] Pierre M.L. Drezet and Robert F. Harrison, “A new sparsity control in support vector classification and regression”, Pattern Recognition 34 (2001) p.111-p.125, 1999
[8] 戴顯權, “資料壓縮”, 紳藍出版社, 2001