Blind Image Deconvolution by Neural Recursive Function Approximation

This work explores blind image deconvolution by recursive function approximation based on supervised learning of neural networks, under the assumption that a degraded image is linear convolution of an original source image through a linear shift-invariant (LSI) blurring matrix. Supervised learning of neural networks of radial basis functions (RBF) is employed to construct an embedded recursive function within a blurring image, try to extract non-deterministic component of an original source image, and use them to estimate hyper parameters of a linear image degradation model. Based on the estimated blurring matrix, reconstruction of an original source image from a blurred image is further resolved by an annealed Hopfield neural network. By numerical simulations, the proposed novel method is shown effective for faithful estimation of an unknown blurring matrix and restoration of an original source image.

• Jiann-Ming Wu
• Hsiao-Chang Chen
• Chun-Chang Wu
• Pei-Hsun Hsu

• Blind image deconvolution
• linear shift-invariant(LSI)
• linear image degradation model
• radial basis functions (rbf)
• recursive function
• annealed Hopfield neural networks.

[1] L. Bar, N. Sochen and N. Kiryati, "Semi-blind image restoration via
Mumford-Shah regularization," IEEE Trans. Image Process., vol. 15, no.
2, pp. 483--493, Feb 2006.
[2] D. Kundur and D. Hatzinakos, "Blind image deconvolution," IEEE Trans.
Signal Process. Mag., vol. 13, no. 3, pp. 43--64, Mar. 1996.
[3] T. E. Bishop, S. D. Babacan, B. Amizic, A. K. Katsaggelos, T. Chan, and
R. Molina, "Blind image deconvolution: Problem formulation and
existing approaches," in Blind Image Deconvolution: Theory and
Applications, P. Campisi and K. Egiazarian, Eds. Boca Raton, FL: CRC,
ch. 1, 2007.
[4] Nan Xie and Henry Leung, Member, "IEEE Blind Equalization Using a
Predictive Radial Basis Function Neural Network," IEEE Trans. Neural
Networks., vol. 16, no. 3, May 2005.
[5] J. M. Wu and C. S. Lian, "Blind Deconvolution by Supervised Learning
of Neural Networks," Master thesis, Department of Applied Mathematics,
National Dong Hwa University, Jul. 2008.
[6] S. Kullback, "Information Theory and Statistics," New York: Dover,
1959.
[7] A. C. Likas and N. P. Galatsanos, "A variational approach for Bayesian
blind image deconvolution," IEEE Trans. Signal Process., vol. 52, no. 8,
pp. 2222--2233, 2004.
[8] K. Z. Adami, "Variational methods in Bayesian deconvolution," in Proc.
PHYSTAT ECONF, 2003, vol. CO30908, p. TUGT002.
[9] S. D. Babacan, R. Molina and A. K. Katsaggelos, "Variational Bayesian
Blind Deconvolution Using a Total Variation Prior," IEEE Trans. Image
Process., vol. 18, no. 1, Jan. 2009.
[10] R. Jirik and T. Taxt, "Two-dimensional blind Bayesian deconvolution of
medical ultrasound images," IEEE Trans. Ultrasonics, Ferroelectrics, and
Frequency Control., vol 55, no 10, pp. 2140 - 2153, Oct. 2008.
[11] D. G. Tzikas, A. C. Likas and N. P. Galatsanos, "Variational Bayesian
Sparse Kernel-Based Blind Image Deconvolution With Student's-t
Priors," IEEE Trans. Image Process., vol. 18, no. 4, pp. 753 - 764, Apr.
2009.
[12] J. M. Wu, "Annealing by Two Sets of Interactive Dynamics," IEEE Trans.
Systems, Man, and Cybernetics, Part B., vol. 34, no. 3, pp. 1519 - 1525.
Jun. 2004.
[13] J. J. Hopfield and D. W. Tank, "Neural computation of decisions in
optimizationproblems," Biol. Cybern., vol. 52, pp. 141, 1985.
[14] J. B. MacQueen, (1967) . "Some Methods for classification and Analysis
of Multivariate Observations" in Proceedings of 5th Berkeley Symposium
on Mathematical Statistics and Probability. 1: 281--297, University of
California Press. MR0214227. Zbl 0214.46201. Retrieved on
2009-04-07.
[15] Gunnar Rätsch, "A (Partial) Documentation of RBF & AdaBoostReg
Software Packages."
[16] M. N├©rgaard, O. Ravn, N. K. Poulsen and L. K. Hansen. "Neural
Networks for Modelling and Control of Dynamic Systems," London,
U.K.:Springer-Verlag, 2000.
[17] Matlab toolbox "Deblurring Images Using the Blind Deconvolution
Algorithm."