臺灣博碩士論文加值系統

在臨床醫學上常碰到低對比度的影像，而其正確判讀是非常重要的。我們提出一個基於奇異值分解之影像演算法來強化低對比之骨折X光片，首先須將目標區域分離，之後再行長條圖等化及奇異值選取以呈現影像。奇異值分解的頻譜特性將予探討，而高頻的部份將如同傅利葉分析中予以強化。我們的方法可提供影像的額外面向，在本文中我們的方法不需額外參數的設定。實驗的結果由十位臨床醫師以視覺類比化量表加以評斷，其得分亦由原先的2.5進步到8.3分，結果顯示此方法有良好的效果且將在骨折影像上有所助益。
我們發展了一個可用於醫學影像之強健的類高通濾波器。我們的演算法類似其他方法均有一旋積核，其數學型式類似變異數統計量且是自我適應性的。我們探索其數學表徵型式並改寫成矩陣的二次式。此運算子具備高度的內在結構及廣泛的空間對稱性。我們的方法不會邊界斷裂、位移及寬度變形，亦不會對雜訊太敏感，因此能有效的將邊緣偵測出來。我們以王、林兩氏來對此運算子命名，它的效能經由專家以視覺類比化量表加以評定，在不同的醫學影像如X光片、電腦斷層及核磁共振上它的表現均在變異數分析上有統計的顯著差異。 因此，此運算子值得進一步去評估。

Low contrast profile images are frequently encountered in medical practice and correct interpretation of these images is of vital importance. We propose a contrast enhancement technique based on singular value decomposition (SVD) to enhance the low contrast fracture X-ray image. The region of interest is manually cropped, then histogram equalization (HE) and singular value selection procedure follows for further image presentation. The spectral property of SVD is exploited and singular value selection technique is developed on the analogy to the Fourier domain technique for high frequency enhancement. Our method can generate extra viewpoints of the target images to supplement the HE processing. The proposed singular value selection technique does not need any arbitrary parameters as for X-ray enhancement in this paper. The performance of our work was justified by ten physicians and presented with Visual Analogue Scale (VAS). The average VAS score improves from 2.5 with HE along to 8.3 by the proposed method. Experimental results indicate that the proposed method has promising result and is helpful in fracture X-ray image processing.
We develop a robust, quasi- highpass filter for edge detection in medical images. Our algorithm is kernel-based one similar to conventional edge detectors. The edge detector we proposed has mathematical form of local variance and is adaptive in nature. The mathematical formulation of the detector is exploited and re-expressed as quadratic form of Toeplitz matrix. The detector has highly structured internal architecture with abundant spatial isotropic symmetricity. With our proposed operator, the frequently encountered problems in edge detection such as fragmentation, position dislocation, and thinness loss are greatly diminished. The detector is robust to noise and has excellent ability to extract the important edge features contained in object boundaries. We named this new operator as WL-operator (Wang and Lin). The performance of WL-operator is compared to the other edge detectors and justified with experts using Visual Analogue Scale (VAS) scores. Results for different medical imaging modalities including X-ray, CT, and MRI are promising, with statistical significance demonstrated by Analysis of Variance (ANOVA). Experimental results indicate that the WL-operator has good performance and is helpful in medical image processing.

Chapter 1. 基於奇異值分解之骨折影像強化…………………………………...……12
I. Introduction……………………………………………………………………..…12
II. Pre-processing and HE technique……………………………………………..….15
A. Cropping the image for ROI…………………………………………………...15
B. Histogram equalization processing…………………………………………….16
III. SVD Selection Technique……………………………………………………….18
A. Fundamental Elements of Singular Value Decomposition…………………….18
B. Spectral Analysis of the SVD principal images……………………………..…21
C. Genetic Algorithm Simulation of HE………………………………………….23
D. Singular Value Selection for Complement Images…………………………….25
E. Stopping criteria for Selection number of Complement Images……………….26
IV. Experimental Results and Discussion……………………………………………27
V. Conclusion………………………………………………………………………...32

Part 2. 類高通濾波器之醫學影像邊緣偵測……………………………………….…61
I. Introduction……………………………………………………………….……….61
II. Edge Detection Operators…..………………………………..…………………...65
A. Previous work…………….……………………………………………………66
B. The proposed edge detector…….………………………………………….…..67
III. Properties of the WL Operator …………………………………… …………….69
A. Quadratic form presentation of the WL operator…………………..…………..69
B. Isotropic symmetricity of the WL pperator……….…………………….……..71
C. Dynamic range of the WL operator………………………………………..…. 72
D. Histogram of the WL operator response in medical images ..………...……….73
E. Correspondence with other edge detectors…………………………………….74
IV. Experimental Results and Discussion……………………………………………76
V. Conclusion………………………………………………………………………...79
References…………………………………………………………………………….103

References
[1] Gonzalez, Woods, Eddins, Digital Image Processing, Prentice-Hall, Englewood Cliffs, NJ, 2004.
[2] Z. Yu, C. Bajaj, “A fast and adaptive method for image contrast enhancement,” Int. Conf. on Image Processing (ICIP), Oct. 2004.
[3] S.D. Chen, A.R. Ramli, “Contrast enhancement using recursive mean-separate histogram equalization for scalable brightness preservation,” IEEE Trans. Consumer Electron., Vol. 49, No. 4, pp.1301-1309, 2003.
[4] J.Y. Kim, L.S. Kim, and S.H. Hwang, “An advanced contrast enhancement using partially overlapped sub-block histogram equalization,“ IEEE Trans. on Circuits and Systems for Video Technology, Vol. 11, pp. 475-484, 2001.
[5] J.A. Stark. “Adaptive contrast enhancement using generalization of histogram equalization,” IEEE Trans. Image Processing, Vol. 9, no. 2, pp. 889-906,2000.
[6] H. Y. Yang, Y. C. Lee, et al. “A novel algorithm of local contrast enhancement for medical image,” IEEE Nuclear Science Symposium Conf. Record, Vol. 5, pp. 3951-3954, 2007.
[7] J. Lu, “Contrast enhancement of medical images using multiscale edge representation,” Proc. SPIE, Vol. 2242, pp. 711, 1994.
[8] P.C. Hansen, “The truncated SVD as a method for regularization,” BIT archive, Vol. 27, Issue 4, pp. 534-553, 1987.
[9] K. Dan, “A singularly valuable decomposition: the SVD of a matrix,” The College Math. J. Vol. 27 No.1, pp. 2-23, Jan 1998.
[10] G.W. Stewart, “On the Early History of the Singular Value Decomposition,” SIAM Review 35(4):551-566, 1993.
[11] H.H. Barrett, W. Swindell, Radiological Imaging, Academic Press, New York, 1981, pp. 410-414.
[12] K. Jain, Fundamentals of digital image processing, Englewood Cliffs, NJ, Prentice-Hall, 1989.
[13] Y. T. Kim, “Contrast enhancement using brightness preserving bi-histogram equalization,” IEEE Trans. Consumer Electron., Vol. 43, No. 1, pp. 1-8, 1997.
[14] C.C. Sun, S.J. Ruan, M.C. Shie, and T.W. Pai, “Dynamic contrast enhancement based on histogram specification,” IEEE Trans. on Consumer Electron., Vol 51, No. 4, pp. 1300-1305,2005.
[15] L.N. Trefethen and David Bau III, Numerical Linear Algebra, Society for Industrial and Applied Mathematics, Philadelphia, PA, USA, 1997.
[16] P.C. Hansen, S.H. Jensen, “FIR filter representation of reduced-rank noise reduction,” IEEE Trans. Signal Proc., Vol. 46, pp. 1737-1741, 1998.
[17] V. C. Klema, “The singular value decomposition: its computation and some applications,” IEEE Trans. On Automatic Control, Vol. 25, pp.164-176, 1980.
[18] A. Ranade, S.S. Mahabalarao, S. Kale, “A variation on SVD based image compression,” Image and Vision computing, Vol. 25, Issue 6, pp. 771-777, 2007.
[19] M.E. Wall, A. Rechtsteiner, L.M. Rocha, A practical approach to microarray data analysis, Springer, Norwell, MA, USA, 2003.
[20] D.E. Goldberg, Genetic Algorithms in Search, Optimization and Learning, Kluwer Academic Publishers, Boston, MA, 1989.
[21] P.S. Myles, S.T. Troedel, M. Boquest, M. Reeves, “The pain visual analog scale: is it linear or nonlinear?”, Anesth. Analg. Vol. 89, pp. 1517-20, 1999.
[22] F. Dexter, D.H. Chestnut, “Analysis of statistical tests to compare visual analog scale measurements among groups,” Anesthesiology Vol. 82, pp. 896-902,1995.
[23] M. Gudmundsson, E.A. El-Kwae, M. R. Kabuka, “Edge detection in medical images using a genetic algorithm”, IEEE Trans. Medical Imaging, Vol. 17, No. 3, June 1998.
[24] M. Basu, “Gaussian-based edge-detection methods a survey”, IEEE Trans. on Systems, Man, and Cybernetics., Col. 32, No. 3, Aug 2002.
[25] D. Xu, T. Kasparis, “Detection and localization of edge contours”, Proc. SPIE, Vol. 5097, 2003.
[26] D. Williams and M. Shah. Edge contours using multiple scales. Comput. Vision Graphics Image Processing. Vol. 51, pp.256-274, 1990.
[27] D. Ziou and S. Tabbone, “Edge detection technique- an overview,” Int. J. Pattern Recognit. Imag Anal., vol. 8, pp. 537-559, 1998.
[28] V.S. Nalwa and T.O. Binford. On detecting edges. IEEE trans Pattern Anal Machine Intell vol PAMI-8, pp. 699-714, June 1986.
[29] L.J. Vliet, I.T. Young, A.L.D. Beckers, “An edge detection model based on non-linear Laplace filtering”, Patten Recog. And Artificial Intell, E.S. Gelsema & L.N. Kanal(eds), Elsevier, 1988, 63-73.
[30] M.J. Black, G. Sapiro, D.H. Marimount, D. Heeger, “Robust anisotropic diffusion”, IEEE Trans. Image Process. Vol. 7, pp. 421-432, 1998.
[31] Z.J. Hou, G.W. Wei, “A new approach to edge detection”, Pattern Recog. Vol 35, pp. 1559-1570, 2002.
[32] M. Basu, “Gaussian derivative model for edge enhancement,” Pattern Recognit., vol. 27, pp. 1451-1461, 1994.
[33] W.K. Pratt, Digital Image Processing. New York: Wiley-interscience, 1978.
[34] V. Torre and T. Poggio, “On edge detection,” IEEE Trans. Pattern Anal. Machine Intell., Vol. PAMI-8, pp. 147-163, 1986.
[35] D. Marr and E. Hildreth, “Theory of edge detection,” Proc. R. Soc. Lond. A, Math. Phys. Sci., vol. B 207, pp. 187-217, 1980.
[36] S. Castan, J. Zhao, J. Shen, “New edge detection methods on exponential filter”, Pattern Recognit., 1990, Proc. Vol.1 pp.709-711.
[37] R.J. Qian and T.S. Huang, “Optimal edge detection,” IEEE Image Trans. Image Processing, Vol. 5, pp. 1215-1220, 1996.
[38] F. Bergholm, “Edge focusing,” IEEE Trans. Pattern Anal. Machine Intell., Vol. PAMI-9, pp. 726-741, 1987.
[39] X. Dai, S. Khorram, “A feature-based image registration algorithm using improved chain-code representation combined with invariant moments,” IEEE Trans. Geosci. Remote Sensing, Vol. 37, pp. 2351-2362, 1999.
[40] A. Yulle and T.A. Poggio, “Scaling theorems for zero crossings, “ IEEE Trans. Pattern Anal. Machine Intell., Vol. PAMI-8, pp. 15-25, 1986.
[41] M.D. Heath, S. Sarkar, T. Sanocki, K.W. Bowyer, “ A robust visual method for assessing the relative performance of edge-detection algorithms”, IEEE Trans. Pattern Anal. Mach. Intell. Vol. 19, pp. 1338-1359, 1997.
[42] I.E. Abdou, W.K. Pratt, “Quantitative design and evaluation of enhancement / thresholding edge detectors”, Proc. IEEE. Vol 69, pp.753-763, 1979.