臺灣博碩士論文加值系統

本論文提出一種快速影像分割的類神經網路，稱之為共識原則之類神經網路(CVNN)，此網路的訓練演算法使用一種稱為共識原則(Consensus Voting)的概念。共識原則的一個重要特性是：目標神經元的輸出標籤(label)是由其對應的相鄰神經元的輸出標籤所共同決定，也就是說鄰居神經元的輸出會影響目標神經元的權重改變，進而決定其標籤值。本質上，CVNN的訓練演算法主要是反覆利用目標神經元與其相鄰神經元之間的相互影響，使彼此形成一種共識(要被貼上何種標籤)，以達到影像分割的目的。由於CVNN的訓練演算法僅使用”選票統計”的策略來進行權重更新，因此其訓練演算法是簡單且快速的。大量的實驗結果證實CVNN的確能夠有效的進行影像分割，同時，含有雜訊的輸入影像也用來展示CVNN的強健性。相較於傳統的分割技術CVNN能達到更有效率的計算與更精確的分割結果。

This paper presents a novel approach, called Consensus Voting Neural Network (CVNN) which aims to perform fast and accurate image segmentation for gray images. To achieve this goal, a pertaining training algorithm based on the principle of vote-to-consensus is developed. The essence of the CVNN approach is the iterative interaction between the target neuron and its neighboring pixels, the range of which is defined by the running mask. Specifically, the neighboring neurons surrounding the target neuron collaboratively determine the label for the target neuron in a way of voting with their respective labels. Due to its simplicity in the updating strategy that solely uses scalar increment in the ballot-counter, the training algorithm of CVNN is fast. Extensive experimental results are provided to verify the effectiveness of CVNN using various benchmark input images. Robustness of CVNN to noise is also illustrated. All comparison results have indicated that the proposed CVNN outperforms CSNN in terms of computation efficiency, robustness, and segmentation accuracy.

CHAPTER 1 2
INTRODUCTION 2
1-1. Review of Image Segmentation 2
1-2. Motivation 6
1-3. Outline of The Thesis 11

CHAPTER 2 12
CONSENSUS VOTING NEURAL NETWORK 12
2-1. The Network Architecture 12
2-2. Principle of Vote-to-Consensus 14
2-3. Training Algorithm for CVNN 15

CHAPTER 3 24
CHARACTERIZATIONS AND SIMULATION RESULTS 24
3-1. Applying CVNN to Image Segmentation 24
3-2. Noise Tolerance in Image Segmentation 33
3-3. Convergence Analysis 40

CHAPTER 4 43
CONCLUSIONS AND DISCUSSIONS 43
REFERENCES 44

[1] N. R. Pal, and S. K. Pal, “A review on image segmentation techniques,” Pattern Recognetion, vol. 26, pp. 1277-1294, 1993.
[2] K. Mardia and T. Hainsworth, “A spatial thresholding method for image segmentation,” IEEE Trans. Pattern Anal. Machine Intell., vol. 10, pp.919-927, Nov. 1988.
[3] O. Wonho and W. B. Lindquist, “Image thresholding by indicator krigin,”IEEE Trans. Pattern Anal. Machine Intell., vol. 21, pp.590-602, Jul. 1999.
[4] K. S. Fu and J. K. Mui, A survey on image segmentation,”Pattern Recognetion, vol. 13, pp. 3-16, 1981.
[5] L. S. Davis, “A survey of edge detection techniques,” comput. Graphics Image Process, vol. 4, pp. 248-270, 1975.
[6] A. Rosenfeld and A. C. Kak, Digital Picture Processing, Academic Press, New York, 1982.
[7] E. L. Hall, Computer Image Processing and Recognition, Academic Press, New York, 1979.
[8] R. C. Gonzalez and P. Wintz, Digital Image Processing, Addison-Wesley, Reading, Massachusetts, 1987.
[9] R. C. Gonzalez and R. E. Woods, Digital Image Processing, Addison-Wesley, Reading, Mass, 1993.
[10] Y. L. Chang and X. Li, Adaptive image region growing,”IEEE Trans. Image Processing, vol. 3, no. 6, pp. 868-873, 1994.
[11] J. P. Beaulieu and M. Goldberg, A new approach to combining region growing and edge detection,”Pattern Recognition Letters, vol. 14, pp. 869-875, 1993.
[12] S. A. Hojjatoleslami and J. Kittler, “Region growing: a new approach,” CVSPP Technical Report TR-6/95, University of Surrey, Department of Electronic and Electrical Engineering, 1995.
[13] A. V. Wangenheim, H. Wagner, D. Krechel, and P. Conrad, “Detecting and correcting failed segmentations of radiological images using a knowledge-based approach,” IEEE Proc. Int. Conf. Computer-Based Medical Systems, pp. 175-180, 2000.
[14] D. Hagyard, M. Razaz, and P. Atkin, “Analysis of watershed algorithms for greyscale images,” IEEE Proc. Int. Conf. Image Processing, pp. 41-44, 1996.
[15] S. Haykin, Neural Networks: A Comprehensive Foundation, 2nd ed. Hamilton Ontario, CA:Prentice-Hall, 1999.
[16] M. L. Minsky, “Steps towards artificial intelligence,” Proceedings of the Institute of Radio Engineers, vol. 49, pp. 8-30, 1961. (Reprinted in: E. A. Feigenbaum, and J. Feldman, eds., Computers and Thought, pp. 406-450, New York: McGraw-Hill.)
[17] M. L. Minsky, and S. Papert, Perceptrons, Cambridge, MA: MIT Press, 1969.
[18] D. E. Rumelgart, G. E. Hinton and R. J. Williams, “Learning internal representations by error propagation.” In Parallel Distributed Processing: Explorations in the Microstructure of Cognition, vol. 1, Chapter 8, Cambridge, MA: MIT Press.
[19] P. S. Churchland, and T. J. Sejnowski, The Computational Brain, Cambridge, MA: MIT Press, 1992.
[20] J. J. Hopfield, “Neural networks and physical systems with emergent collective computational abilities,” Proceeding of the National Academy of Sciences, USA, vol.79, pp. 2554-2558, 1982.
[21] J. J. Hopfield, and D. W. Tank, “‘Neural’ computation of decisions in optimization problems,” Biological Cybernetics, vol. 52, pp. 141-152, 1985.
[22] J. M. Mendel and R. W. McLaren, “Reinforcement-learning control and pattern recognition systems,” Adaptive, Learning, and Pattern Recognition Systems: Theory and Applications, vol. 66, J.M. Mendel and K.S. Fu, eds., pp.287-318, New York: Academic Press, 1970.
[23] N. Papamarkos, C. Strouthopoulos, and I. Andreadis, “Multithresholding of colour and gray-level images through a neural networks,” Image and Vision Computing, vol. 18, pp. 213-222, 2000.
[24] M. W. Ho, “Segmentation of MRI Using Fuzzy Clustering and Multiresolution Techniques,” MS. Thesis, Nuclear Science Department, NTHU, 2002.
[25] N. Babaguchi, K. Yamada, K. Kise, and T.Tezuka, “Connectionist model binarization,” Proc. 10th ICPR, pp. 51-56, 1990.
[26] Y. B. Yu, “The Study of License Plate Numbers Recognition Using Plastic Perceptron Neural Network,” MS. Thesis, Computer and Communication Engineer Department, NKFUST, 2002.
[27] M. P. Chen, “The Applications of Neural Networks on Medical Images,” MS. Thesis, Computer Science and Information Engineering Department, THU, 2000.
[28] C. T. Chen, E. C. Tsao, and W. C. Lin, “Constraint satisfaction neural networks for image segmentation,” Pattern Recognition, vol. 25, no. 7, pp. 679-693, 1991.
[29] F. Kurugollu, B. Sankur, and A. Emre Harmanci, “Image Segmentation by Relaxation using Constraint Satisfaction Neural Network,” Image and Vision Computing, vol. 20, pp. 483-497, 2002.
[30] B. Kosko, Neural Networks and Fuzzy Systems: A Dynamical Systems Approach. Upper Saddle River, NJ, Prentice-Hall, 1991.
[31] B. Widrow, and M. E. Hoff, Jr., “Adaptive switching circuits,” IRE WESCON Convention Record, pp. 96-104.
[32] B. Fritzke, “Growing Cell Structures – A Self-organizing Network for Unsupervised and Supervised Learning,” Neural Networks, vol.7 pp. 1441-1460, 1994.
[33] T. Kohonen, “Self-organized formation of topologically correct feature maps,” Biol. Cybern. vol. 43, pp. 59-69, 1982.
[34] J. W. Goodman, “Some fundamental properties of speckle,” J. Opt. Soc. Am., vol. 66, no. 11, Nov. 1976.