旋轉移動模糊影像還原之研究

Research on Restoration of Images Blurred by Circular Motion

侯谷學(Ku-Hsueh Hou)

指導教授 ： 林啟芳

元智大學/資訊學院/資訊工程學系/碩士(2013年)

https://doi.org/10.6838/YZU.2013.00334

摘要

在移動模糊影像處理中，目前研究多偏向於平面ㄧ致性與線性的模糊疊加，對於空間可變的模糊情形涉及較少，尤其是對於旋轉移動模糊。本文比較了幾個常用的影像還原方法，並在電腦模擬旋轉移動模糊影像及數位拍攝裝置拍攝之真實旋轉移動模糊影像上測試影像還原效果，主要目的是為了能夠達到將真實拍攝旋轉移動模糊影像的影像去模糊結果最佳化。在旋轉移動模糊還原的主要工作及研究成果有(1)分別利用空間域及頻率域的電腦模擬方法模擬旋轉移動模糊影像；(2)實驗比較幾種常用及針對性高的影像還原方法，找出其中還原效果最佳的影像還原方法；(3)在估算旋轉移動模糊參數上利用Hough方法找出旋轉中心以及自相關法尋找旋轉角度後，再利用對角梯度加載濾波法還原電腦模擬旋轉移動模糊影像後檢視其還原品質；(4)針對數位影像拍攝裝置所取得的真實旋轉移動模糊影像進行影像還原後檢視其還原品質，並分析還原真實拍攝模糊影像與電腦模擬影像間的差異。

關鍵字

旋轉移動 ； 旋轉移動模糊 ； 影像還原 ； 旋轉角度估算 ； 圓心偵測

並列摘要

Most researches in motion blur image process are related to plane consistence and linear blur stack, yet seldom involve variable blur in space, especially in rational motion blur. An aim of this research was to optimize the result of eliminating blur for blurred images taken under rational motion situation. The rational blur images generated using computer modeling and digital photographic apparatus were recovered with some existing methods, the results of which are compared in the research. The main work and findings for rational blur images are as follows: (1) Generating rational blur images with computer simulation of space and frequency；(2) Finding the optimized method for photo recovery；(3) Recovering the blur images by Diagonal Gradient Loading Filter linking up with rotation parameters estimated with Hough method and Autocorrelation method；(4) Examining the quality of rotational motion blur images taken by digital photographic apparatus after recovery.

並列關鍵字

circular motion ； circular motion blur ； rotation center detection ； rotation angle detection

參考文獻

[2] Sr. J. L. Harris, “Image Evaluation and Restoration,” Journal of the Optical Society of America, vol. 56, no. 5, pp. 569-570, 1966.

[3] C. W. Helstrom, ”Image Restoration by the Method of Least Square,” Journal of the Optical Society of America, vol. 57, no. 3, pp. 297-303, 1967.

[5] B. R. Hunt, “A Matrix Theory Proof of the Discrete Convolution Theorem,” IEEE Transactions on Audio and Electroacoustics, vol. 19, no. 4, pp. 285-288, Dec. 1971.

[6] B. R. Hunt, “The Application of Constrained Least Squares Estimation to Image Restoration by Digital Computer,” IEEE Transactions on Computers, vol. C-22, no. 9, pp. 805-812, Sep. 1973.

[7] X. Mestre and M. #westeur002#. Lagunas, “Finite Sample Size Effect on Minimum Variance Beamformers: Optimum Diagonal Loading Factor for Large Arrays,” IEEE Transactions on Signal Processing, vol. 54, no. 1, pp. 69-82, Jan. 2006.