植基於類神經網路，本論文提出一套針對在數位時間延遲積分馬賽克影像的去馬賽克演算法，而數位時間延遲積分馬賽克影像其主要使用於印刷工業之上。在所提出的方法中，首先，透過異質性映射法決定內插時的資料相依性。之後，再將所決定的資料相依性結合徑向基底類神經網路訓練濾波器，發展出一套高品質的去馬賽克演算法。透過數張常用的測試影像，實驗結果顯示，比較近年來所提出之五種演算法，本論文所提出之演算法在彩色訊號雜訊比與S-CIELAB 兩個主要彩色影像量度上，都能得到較佳的輸出影像品質。

Based on neural network, this paper presents a novel demosaicing algorithm for digital time delay and integration (DTDI) mosaic images, which are suitable for industrial print inspection. According to the adaptive heterogeneity projection masks, the data dependence of the interpolation estimation can be first determined. Combining the determined data dependence and radial-basis function (RBF) neural network, a quality-efficient demosaicing algorithm for DTDI mosaic images is developed. Based on some popular test images, the proposed demosaicing algorithm has better image quality performance in terms of the color peak signal-to-noise ratio (CPSNR) and the S-CIELAB when compared with several previous demosaicing algorithms.

[1]R. Lukac and K. N. Plataniotis, “Color filter arrays: Design and performance analysis,” IEEE Trans. Consum. Electron., vol. 51, no.4, pp. 1260–1267, 2005.
[2]B. E. Bayer, “Color imaging array,” U.S. Patent, no. 3971 065, 1976.
[3]R. Huber-Mörk, H. Ramoser, H. Penz, K. Mayer, D. Heiss-Czedik, and A. Vrabl, “Region based matching for print process identification,” Pattern Recognit. Lett., vol. 28, no. 15, pp. 2037–2045, 2007.
[4]J. Fürtler, E. Bodenstorfer, K. J. Mayer, J. Brodersen, D. Heiss-Czedik, H. Penz, C. Eckel, K. Gravogel, and H. Nachtnebel, “High performance camera module for fast quality inspection in industrial printing applications,” in Proc. of Electronic Imaging Conf. on Machine Vision Applications in Industrial Inspection, vol. 6503, pp. 65030J-1–65030J-12, 2007.
[5]E. Bodenstorfer, J. Fürtler, J. Brodersen, K. J. Mayer, C. Eckel, K. Gravogel, and H. Nachtnebel, “High speed line-scan camera with digital time delay integration,” in Proc. of Electronic Imaging Conf. on Real Time Imaging, vol. 6496, pp. 64960I-1–64960I-10, 2007.
[6]T. Sakamoto, C. Nakanishi, and T. Hase, “Software pixel interpolation for digital still camera suitable for a 32-bit MCU,” IEEE Trans. Consum. Electron., vol. 44, no. 4, pp. 1342–1352, 1998.
[7]C. Larocheand and M. Prescott, “Apparatus and method for adaptively interpolating a full color image utilizing chrominance gradients,” US Patent, no. 5, pp. 373–322, 1994.
[8]D. Heiss–Czedik, R. Huber–Mörk, D. Soukup, H. Penz, and B. L. Garc’ıa, “Demosaicing algorithms for area- and line-scan cameras in print inspection,” J. Vis. Commun. Image Represent., vol. 20, no. 6, pp. 389–398, 2009.
[9]D.S. Broomhead and D. Lowe, “Multivariable functional interpolation and adaptive networks,” Complex Systems, vol. 2, no. 3, pp. 321-355, 1988.
[10]J. E. Moody and C. Darken, “Fast learning in networks of locally-tuned processing units,” Neural Computation, vol. 1, no. 2, pp. 281-294, 1989.
[11]T. Poggio and F. Girosi, “Regularization algorithms for learning that are equivalent to multilayer networks,” Science, vol. 247, no. 4945, pp. 978-982, 1990.
[12]K. L. Chung, W. J. Yang, W. M. Yan, and C. C. Wang, “Demosaicing of color filter array captured images using gradient edge detection masks and adaptive heterogeneity-projection,” IEEE Trans. Image Processing, vol. 17, no. 12, pp. 2356-2367, 2008.
[13]K. L. Chung, W.J. Yang, P.Y. Chen, W.M. Yan, and C.S. Fuh, “ New joint demosaicing and zooming algorithm for color filter array,” IEEE Trans. Consumer Electronics, vol. 55, no. 3, pp.1477-1486, 2009.
[14]K. L. Chung, W.J. Yang, J.H. Yu, W.M. Yan, and C.S. Fuh, “Novel quality-effective zooming algorithm for color filter array,” Journal of Electronic Imaging, vol. 19, no. 1, pp. 013005-1-013005-15, 2009.
[15]S. C. Pei and I. K. Tam, “Effective color interpolation in CCD color filter arrays using signal correlation,” IEEE Trans. Circuits and Systems for Video Technology, vol. 13, no. 6, pp. 503-513, 2003.
[16]D. E. Rumelhart, G. E. Hinton and R. J. Williams, “Learning internal representations by error propagation,” Parallel Distributed Processing, vol. 1, pp. 318–362, 1986.
[17]M. Riedmiller and H. Braun, “A direct adaptive method for faster back propagation learning: the RPROP algorithm,” in: Proc. of the International Conf. on Neural Network, vol. 1, pp.586-591, 1993.
[18][Online]. Available: http://www.site.uottawa.ca/~edubois/demosaicking/.
[19]R. W. G. Hunt, Measuring Colour, 2nd ed. Chichester, U.K.: Ellis Horwood, 1995.
[20]W. Lu and Y. P. Tan, “Color filter array demosaicking: New method and performance measures,” IEEE Trans. Image Process., vol. 12, no. 10, pp. 1194–1210, 2003.