Title :
Blind image restoration based on total variation regularization of blurred images
Author :
Ohkoshi, Kyosuke ; Goto, Tomio ; Hirano, Satoshi ; Sakurai, Masaru
Author_Institution :
Dept. of Comput. Sci. & Eng., Nagoya Inst. of Technol., Nagoya, Japan
Abstract :
In this paper, we propose a novel blind image restoration method based on total variation (TV) regularization. It involves alternate iteration of point spread function (PSF) estimation and deconvolution. Using this method, we can obtain clear images from blurred images without an increase in noise and ringing. Thus, blurred images captured using digital cameras can be restored effectively.
Keywords :
deconvolution; image restoration; iterative methods; blind image restoration; blurred image; deconvolution iteration; digital camera; point spread function estimation iteration; total variation regularization; Deconvolution; Equations; Estimation; Image restoration; Kernel; Mathematical model; Noise; Deblurring; Image restoration; Motion blur;
Conference_Titel :
Consumer Electronics (GCCE), 2012 IEEE 1st Global Conference on
Conference_Location :
Tokyo
Print_ISBN :
978-1-4673-1500-5
DOI :
10.1109/GCCE.2012.6379933