簡易檢索 / 詳目顯示

回結果列表
研究生： 呂全斌
Chuan-Pin Lu
論文名稱： 數位訊號分離與圖形理解之研究
A Study of Digital Signal Separation and Pattern Recognition
指導教授： 邱士軒
Shih-Hsuan Chiu
口試委員: 溫哲彥
Che-Yen Wen
邱顯堂
Hsien-Tang Chiu
黃昌群
Chang-Chiun Huang
李俊毅
Jiunn-Yih Lee
何明果
Ming-Guo Her
康淵
Yuan Kang
學位類別： 博士
Doctor
系所名稱： 工程學院 - 材料科學與工程系
Department of Materials Science and Engineering
論文出版年： 2006
畢業學年度： 94
語文別： 英文
論文頁數： 107
中文關鍵詞： 數位影像處理 、訊號分離 、圖形識別 、數位訊號處理
外文關鍵詞： digital image processing, signal separation, pattern recognition, digital signal processing
相關次數： 點閱：269  下載：4 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 對於分離混合的訊號或多變數資料而言，ICA (independent component analysis)是一個相當有用的技術，並且可以在沒有任何先前的資訊下，找出隨機變數或混合訊號裡隱藏的關係元素，在1994年ICA的問題第一次被提出，之後經過十年的發展才逐漸定義出明確的數學模式，然而至今ICA的理論尚有一些發展的空間(如演算法的複雜性與相關的應用)，在本論文中我們提出改善ICA的方法，並應用ICA來進行數位訊號分離與圖形理解相關研究。

    首先，方法上我們提出一降低演算資料量的演算法與一幾何ICA對混合影像訊號進行分離，於降低演算資料量演算法上，我們以fixed-point ICA (FastICA)為基礎進行性能評估，於幾何ICA的方法上，我們則提出以幾何轉換的手法來替代冗長的迭代運算，快速且正確的找出獨立元素。

    接著，使用ICA來進行實際應用研究，我們將ICA設計成雜訊濾波器，藉由ICA分離元素的特性，成功將混合在張力訊號上的雜訊去除，並遠遠超越了傳統雜訊濾波器所能得到的效果。

    其次，我們提出一異常張力訊號辨識系統，系統中我們同樣的應用了ICA來進行去除雜訊的前處理，並且透過新式的人工智慧技術－SSVM (smooth support vector machines)與特徵分析技術來進行張力訊號的識別。

    最後，我們也進行了影像訊號識別的相關研究，研究中所提出的移動物體分析技術是以三維空間的觀念來分析視訊影像，採用即時偵測移動物體的方法與同時調整攝影機鏡頭倍率的方法，來改善傳統的CCTV (closed-circuit television)保全系統影像品質不適當的問題。

    Independent component analysis (ICA) is a useful technique for extracting signal components or multivariate statistical data without a prior knowledge. The general ICA framework was first clearly stated in 1994. Its mathematical model has been developed after then. However, ICA still has some problems in practical applications, such as tedious computations. In this thesis, we improve ICA and apply it to digital signal separation and pattern recognition.

    First, we propose a data reducing algorithm for improving the computed speed and a novel geometric based ICA for improving the tedious computation of traditional ICA. We use a geometric transform to replace iterative computation.

    Secondly, we carry out a practical application by ICA for the noise removing. We design a noise filter based upon ICA. In the noise removing strategy of the ICA filter, we use the “separating” process to replace the “attenuating” operation.

    Thirdly, we propose an on-line unusual tension recognition system for the twister. The study involves pattern recognition and interpretation of the digital signal for one-dimension tension signal. We use ICA to remove noise before feature extraction, and we also apply the robust smooth support vector machines (SSVM) to recognize the tension signal.

    Lastly, we propose a novel framework and a moving objects detection technique for improving the process of image acquisition in the traditional closed-circuit television (CCTV) security systems. In this study, we also apply ICA to noise removing for improving image quality. This framework follows the scientific working group on imaging technology (SWGIT) guideline and can be used in practical CCTV security systems.

    誌謝 I 摘要 III ABSTRACT IV NOTATION V CONTENTS X FIG. & TABLE INDEX XIII Chapter 1. INTRODUCTION 1 1.1. BLIND SOURCE SEPARATION 2 1.2. SURVEY OF INDEPENDENT COMPONENT ANALYSIS 4 1.3. INDEPENDENT COMPONENT ANALYSIS DEFINITION AND APPLICATIONS 6 1.4. PATTERN RECOGNITION OF DIGITAL SIGNAL. 8 1.5. OBJECTIVE AND STRATEGIES OF THIS STUDY 9 Chapter 2. BASIC THEORY OF INDEPENDENT COMPONENT ANALYSIS 10 2.1. NEURAL NETWORKS BASED INDEPENDENT COMPONENT ANALYSIS 11 2.2. FIXED-POINT INDEPENDENT COMPONENT ANALYSIS (FASTICA) 14 2.2.1. Contrast function for ICA 14 2.2.2. Fixed-point algorithm for ICA 17 2.3. SCATTER DIAGRAM BASED INDEPENDENT COMPONENT ANALYSIS 19 Chapter 3. A HISTOGRAM BASED DATA - REDUCING ALGORITHM OF ICA 22 3.1. INTRODUCTION 23 3.2. THE DATA-REDUCING ALGORITHM 24 3.2.1. The coarse step 26 3.2.2. The fine step 27 3.2.3. An example 28 3.3. EXPERIMENTAL RESULTS AND COMPARISONS 30 3.4. CONCLUSIONS 34 Chapter 4. INDEPENDENT COMPONENT ANALYSIS BY GEOMETRIC CORRECTION OF THE SCATTER DIAGRAM 35 4.1. INTRODUCTION 36 4.2. THE PROPOSED METHOD (GEOMETRIC ICA) 37 4.2.1. Translation compensation 39 4.2.2. Correction data points addition 42 4.2.3. Whitening process 43 4.2.4. Geometric rotation 45 4.2.5. Slant compensation 45 4.3. EXPERIMENTAL RESULTS 47 4.4. CONCLUSIONS 54 Chapter 5. NOISE SEPARATION OF THE YARN TENSION SIGNAL ON TWISTER USING FASTICA 55 5.1. INTRODUCTION 56 5.2. THEORY OF FASTICA 59 5.3. EXPERIMENTAL RESULTS 61 5.4. CONCLUSION 69 Chapter 6. UNUSUAL TENSION SIGNAL RECOGNITION BY SMOOTH SUPPORT VECTOR MACHINE 70 6.1. INTRODUCTION 71 6.2. UNUSUAL EVENT CHECK 73 6.3. FEATURE ANALYSIS 75 6.4. SMOOTH SUPPORT VECTOR MACHINE (SSVM) 77 6.5. EXPERIMENTAL RESULTS 79 6.6. CONCLUSIONS 83 Chapter 7. A MOTION DETECTION BASED FRAMEWORK FOR IMPROVING IMAGE QUALITY OF CCTV SECURITY SYSTEMS 84 7.1. INTRODUCTION 85 7.2. METHODS 87 7.2.1. Camera position translation analysis 88 7.2.2. Moving-object detection 90 7.2.3. Zoom control algorithm 93 7.3. EXPERIMENTS 94 7.4. CONCLUSIONS 99 CONCLUSIONS 100 REFERENCES 102

    1.Agbinya, J. I. and D. Rees, “Multi-object tracking in video,” Real-Time Imaging;Vol. 5, No. 5, pp. 295–304 (1999).
    2.Bernd, H, S. Thomas, P. Sam, and P. Tomaso, “Hierarchical classification and feature reduction for fast face detection with support vector machines,” Pattern Recognition, Vol. 36, pp. 2007-2017 (2003).
    3.Burel, G., “Blind separation of sources: A nonlinear neural algorithm,” Neural Network, Vol. 5, pp. 937-947 (1992).
    4.Brown, M., S. R. Gunn, and H. G. Lewis, “Support vector machines for optimal classification and spectral unmixing,” Ecological Modelling, Vol. 120, pp. 167-179 (1999).
    5.Cardoso, J. F., “Blind signal separation: statistical principles,” Proceedings of the IEEE 86, pp. 2009-2025 (1998).
    6.Chen, L. H., Y. C. Lai, C. W. Su, and H. Y. Mark Liao, “Extraction of video object with complex motion,” Pattern Recognition Letters; Vol. 25, No. 11, pp. 1285-1291 (2004).
    7.Cichocki, A., T. Rutkowski, and K. Siwek, “Blind signal extraction of signals with specified frequency band,” Proceedings of the 2002 12th IEEE Workshop on Neural Networks for Signal Processing, pp. 515-524 (2002).
    8.Cristianini, N. and J. Shawe-Taylor, An Introduction to Support Vector Machines and Other Kernal-Based Learning Methods, Cameridge University Press, New York, pp. 149-161 (2000).
    9.Duda, R. O., P. E. Hart, and D. G. Stork, Pattern Classification, A Wiley-Interscience Publication, New York, pp 288-293 (2001).
    10.Geiger, D., A. Gupta, L. A. Costa, and J. Vlontzos, “Dynamic programming for detecting, tracking, and matching deformable contours,” IEEE Trans Pattern Anal Machine Intell, Vol. 17, pp. 294-302 (1995).
    11.Girolami, M. and C. Fyfe, “Generalised independent component analysis through unsupervised learning with emergent Bussgang properties,” In: International Conference on Neural Networks, Vol. 3, pp. 1788-1791 (1997).
    12.Gonzalez, R. C. and R. E. Woods, Digital Image Processing 2nd. Prentice-Hall Publication, New Jersey, pp 655-659 (2002).
    13.Gordon, D., E. Robertson, and J. J. Dowling, “Design and responses of Butterworth and critically damped digital filters,” Journal of Electromyography and Kinesiology, Vol. 13, pp. 569-573 (2003).
    14.Hyvärinen, A. and E. Oja, “A fast fixed-point algorithm for independent component analysis,” Neural Computation, Vol. 9, No. 7, pp. 1483-1492 (1997).
    15.Hyvärinen, A., P. Hoyer, and E. Oja, “Sparse code shrinkage for image denoising,” Neural Networks Proceedings on IEEE World Congress on Computational Intelligence, Vol. 2, pp. 859-864 (1998).
    16.Hyvärinen, A., “Noisy independent component analysis, maximum likelihood estimation, and competitive learning,” Neural Networks Proceedings on IEEE World Congress on Computational Intelligence, Vol. 3, pp. 2282-2287 (1998).
    17.Hyvärinen, A., “New approximations of differential entropy for independent component analysis and projection pursuit,” in Advances Neural Information Processing System, Vol. 10, pp. 273-279 (1998).
    18.Hyvärinen, A., “Fast and robust fixed-point algorithms for independent component analysis,” IEEE Transactions on Neural Networks, Vol. 10, pp. 626-634 (1999).
    19.Hyvärinen, A. and E. Oja, “Independent component analysis: algorithms and applications,” Neural Networks, Vol. 13, pp. 411-430 (2000).
    20.Hyv&auml;rinen, A., The fixed-point ICA (FastICA) MATLAB package. Available from <http://www.cis.hut.fi/projects/ica/fastica/>.
    21.Jenssen, R. and T. Eltoft, “Independent component analysis for texture segmentation,” Pattern Recognition, Vol. 36, pp. 2301-2315 (2003).
    22.Jain, R., R. Kasturi, and B. G. Schunck, Machine Vision, McGraw-Hill, New York (1995).
    23.Karhunen, J., E. Oja, L. Wang, R. Vigario, and J. Joutsensalo, “A class of neural networks for independent component analysis,” IEEE Transactions on Neural Networks, Vol. 8, pp. 486-504 (1997).
    24.Khandetsky, V. and I. Antonyuk, “Signal processing in defect detection using back-propagation neural networks,” NDT and E International, Vol. 35, pp. 483-488 (2002).
    25.Kim, J. B. and H. J. Kim, “Efficient region-based motion segmentation for a video monitoring system,” Pattern Recognition Letters, Vol. 24, pp. 113-128 (2003).
    26.Kotropoulos, C. and I. Pitas, “Segmentation of ultrasonic images using Support Vector Machines,” Pattern Recognition Letters, Vol. 24, pp. 715-727 (2003).
    27.Lee, J. H., H. Y. Jung, T. W. Lee, and S. Y. Lee, “Speech enhancement with map estimation and ICA-based speech features,” Electronics Letters, Vol. 36, pp. 1506-1507 (2000).
    28.Lee, Y. J. and O. L. Mangasarian, “SSVM: a smooth support vector machine for classification,” Computational Optimization and Applications, Vol. 20, pp. 5-22 (2001).
    29.Lee, T. W., M. S. Lewicki, and T. J. Sejnowski, “ICA mixture models for unsupervised classification of non-Gaussian classes and automatic context switching in blind signal separation,” IEEE Transactions on Pattern Recognition and Machine Intelligence, Vol. 22, pp. 1078-1089 (2000).
    30.Lendasse, A., G. Simon, V. Wertz, and M. Verleysen, “Fast bootstrap methodology for regression model selection,” Neurocomputing, Vol. 64, pp. 161-181 (2005). SVM
    31.Li, J., “An empirical comparison between SVMs and ANNs for speech recognition,” The First instructional Conference on Machine Learning (iCML-2003),( http://www.cs.rutgers.edu/~mlittman/courses/ml03/iCML03/papers/li.pdf) (2003).
    32.Liang, Y., E. P. Simoncelli, and Z. Lei, “Color channels decorrelation by ICA transformation in the wavelet domain for color texture analysis and synthesis,” In: Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, Vol. 1, pp. 606-611 (2000).
    33.Lin, C. J., “Asymptotic convergence of an SMO algorithm without any assumptions,” IEEE Transactions on Neural Networks, Vol. 13, pp. 248–250 (2002).
    34.Liu, Y., W. Liu, and Y. Zhang, “Inspection of defects in optical fibers based on back-propagation neural networks,” Optics Communications, Vol. 198, pp. 369-378 (2001).
    35.Odobez, J. M. and P. Bouthemy, “Direct incremental model-based image motion segmentation for video analysis,” Signal Processing, Vol. 66, pp. 143-155 (1998).
    36.Mutihac, R. and M. M. Van-Hulle, “Neural network implementations of independent component analysis,” In: Proceedings of Neural Networks for Signal Processing, the 2002 12th IEEE Workshop on 4-6 Sept., pp. 505-514 (2002).
    37.Park, H. M., H. Y. Jung, T. W. Lee, and S. Y. Lee, “Subband-based blind signal separation for noisy speech recognition,” Electronics Letters, Vol. 35, pp. 2011-2012 (1999).
    38.Pal, M. and A. Goel, “Prediction of the end-depth ratio and discharge in semi-circular and circular shaped channels using support vector machines,” Flow Measurement and Instrumentation, Vol. 17, pp. 49-57 (2006).
    39.Robertson, D., E. Gordon, and J. J. Dowling, “Design and responses of Butterworth and critically damped digital filters”, Journal of Electromyography and Kinesiology, Vol. 13, pp. 569-573 (2003).
    40.Shalvi, O. and E. Weinstein, “New criteria for blind deconvolution of nonminimum phase systems (channels),” IEEE Transactions on Information Theory, Vol. 36, No.2, pp. 312-321 (1990).
    41.Shih, P. and C. Liu, “Face detection using discriminating feature analysis and Support Vector Machine,” Pattern Recognition, Vol. 39, pp. 260-276 (2006).
    42.Southard, T. E. and K. A. Southard, “Performance of filters for noise reduction in maxillary alveolar bone imaging,” IEEE Transactions on Biomedical Engineering, Vol. 42, pp. 13-20 (1995).
    43.Southard, T. E. and K. A. Southard, “Design and responses of Butterworth and critically damped digital filters,” Journal of Electromyography and Kinesiology, Vol. 13, pp. 569-573 (2003).
    44.Tailor, D.R., L. H. Finkel, and G. Buchsbaum, “Color-opponent receptive fields derived from independent component analysis of natural images,” Vision Research, Vol. 40, pp. 2671-2676 (2000).
    45.Tu, T. M., P. S. Huang, and P. Y. Chen, “Blind separation of spectral signatures in hyperspectral imagery,” IEE Proceedings on Vision Image and Signal Processing, Vol. 148, pp. 217-226 (2001).
    46.Vapnik, V. N., “An overview of statistical learning theory,” IEEE Transactions on Neural Networks, Vol. 10, pp. 988-999 (1999).
    47.Vigario, R., J. Sarela, and E. Oja, “Independent component analysis in wave decomposition of auditory evoked fields,” Proceedings of ICANN’98 Conference, Artificial Neural Networks, pp. 287-292 (1998).
    48.Yamaguchi, T., K. Hirokawa, and K. Itoh, “Independent component analysis by transforming a scatter diagram of mixtures of signals,” Optics Communications, Vol. 173, pp. 107-114 (2000).
    49.Zhukov, L., D. Weinstein, and C. Johnson, “Independent component analysis for EEG source localization,” IEEE Engineering in Medicine and Biology Magazine, Vol. 19, pp. 87-96 (2000).
    50.Zhan, Y. and D. Shen, “Design efficient support vector machine for fast classification,” Pattern Recognition, Vol. 38, pp. 157-161 (2005).
    51.Scientific Working Group on Imaging Technology (SWGIT). Recommendations and guidelines for using closed-circuit television security systems in commercial Institutions. Forensic Science Communications, Vol. 7(1) 2005.
    52.Scientific Working Group on Imaging Technology (SWGIT). Recommendations and guidelines for the use of digital image processing in the criminal justice system. Forensic Science Communications, Vol. 5(1) (2003).
    53.Scientific Working Group on Imaging Technology (SWGIT), Best practices for document image enhancement, Forensic Science Communications, Vol. 9(3) (2005).
    54.QAI – Yarn Quality Control System User Guide, YU HWA, pp. 25-45 (2003).
    55."Yarn Tension Monitor System User Guide”, Barmag.

    QR CODE