近年來由於數位攝影器材市場的快速發展，如何將一張因為線性運動所造成模糊的數位影像做復原成了一個熱門的話題。在這篇論文□，我們提出了一個針對線性運動模糊影像復原及改善的方法。

整套方法是由兩大部份所組成：模糊函式的偵測與建構和模糊影像的復原。第一部份又可分為三個步驟，首先利用edge direction distribution來對影像模糊的方向做初步的預估，再把auto-correlation模組套入來找出在該方向所對應的模糊量度，最後則採用了total variation的概念來自動調校這兩個參數。而在第二部份，我們把Hamming window的想法加到了inverse filter上以避免在頻堿轉換時所造成的誤差，並把模糊影像和偵測出的模糊函式輸入以得到一張清晰的影像。

在實驗結果中，我們針對不同虛擬及真實的模糊影像來做準確度、復原效果及容錯度來評比，數據都比其它針對線性運動模糊的方法要來的正確。

Restoration of images corrupt by motion blurs has been strongly demanded due to the emergent digital camera market in recent years. In this paper, we propose a novel algorithm for blindly restoring images with uniform motion blur. The proposed algorithm consists of three steps to recover the motion blur parameters. At first, we estimate an approximate motion blur direction from the edge direction distribution of the blurred image. Subsequently, the motion blur extent is estimated by using the autocorrelation method. Finally, we refine the motion blur estimation by using total variation optimization. Experimental results show superior results by using the proposed algorithm over some previous methods for restoring images with uniform motion blur.

[1] M. R. Banham and A. K. Katsaggelos, “Digital image restoration,” IEEE Signal Processing Magazine, Vol. 14, No. 2, pp. 24-41, March 1997.
[2] D. Kundur and D. Hatzinakos, “Blind image deconvolution”, IEEE Signal Processing Magazine, Vol. 13, No. 3, pp.43-64, May 1996.
[3] W.-G. Chen, N. Nandhakumar, W. N. Martin, “Image motion estimation from motion smear – a new computational model.” PAMI 18, pp.412-425, 1996.
[4] R. G. Lane and R. H. T. Bates, “Automatic multidimensional deconvolution,” J. Opt Soc Am A, Vol. 4(1), pp.180-188, 1987.
[5] C. Mayntx, T. Aach and D. Kunz, “Blur identification using a spectral inertia tensor and spectral zeros”, ICIP 99, Vol. 2, pp.24-18, October, 1999.
[6] A. M. Tekalp and H. Kaufman, “On statistical identification of a linear space-invariant image blurs using non-minimum-phase ARMA models”, IEEE Trans Acoust, Speech, Signal Processing, Vol. 36, pp.1360-1363, August 1988.
[7] A. M. Tekalp, H. Kaufman and J. W. Woods, “Identification of image and blur parameters for the restoration of noncausal blurs”, IEEE Trans Acoust, Speech, Signal Processing, Vol. 34(4), pp.963-972, August 1986.
[8] S. J. Reeves and R. M. Mersereau, “Blur identification by the method of generalized cross-validation”, IEEE Trans Image Processing, vol. 1(3), pp. 301-311, July 1992.
[9] V. Z. Mesarovic, N. P. Galatsanos and M. N. Wernick, “Restoration from partially-known blur suing an expectation-maximization algorithm”, IEEE Conference on Signals, Systems and Computers, Vol.1, pp.95-99, November 1996.
[10] G. R. Ayers and J. C. Dainty, “Iterative blind deconvolution method and its applications“, Optics Letters, vol. 13(7), pp.547-549, July 1988.
[11] T. F. Chan. and C.-K. Wang. “Total variational blind de-convolution”, IEEE Trans Image Processing, vol. 7, no. 3, 1998.
[12] Y. Yitzhaky and N. S. Kopeika, “Identification of blur parameters for motion blurred images”, Graphical Models and Image Processing, vol. 59, no. 5, pp. 310-320, 1997.
[13] Moshe. B.-E. and S. K. Nayer, “Motion-Based Motion Deblurring”, IEEE trans on Pattern Analysis and Machine Intelligence, Vol. 26, No. 6, pp. 689-698, June 2004.