Title :
Regularization of RIF blind image deconvolution
Author :
Ng, Michael K. ; Plemmons, Robert J. ; Sanzheng Qiao
Author_Institution :
Dept. of Math., Hong Kong Univ., Hong Kong
fDate :
6/1/2000 12:00:00 AM
Abstract :
Blind image restoration is the process of estimating both the true image and the blur from the degraded image, using only partial information about degradation sources and the imaging system. Our main interest concerns optical image enhancement, where the degradation often involves a convolution process. We provide a method to incorporate truncated eigenvalue and total variation regularization into a nonlinear recursive inverse filter (RIF) blind deconvolution scheme first proposed by Kundar, and by Kundur and Hatzinakos (1996, 1998). Tests are reported on simulated and optical imaging problems
Keywords :
convolution; deconvolution; image enhancement; image restoration; inverse problems; nonlinear filters; optical information processing; recursive filters; RIF blind image deconvolution; blind image restoration; blur; convolution process; degradation sources; degraded image; imaging system; nonlinear recursive inverse filter blind deconvolution scheme; optical image enhancement; optical imaging; partial information; regularization; total variation regularization; true image; truncated eigenvalue; Convolution; Deconvolution; Degradation; Eigenvalues and eigenfunctions; Image enhancement; Image restoration; Nonlinear optics; Optical filters; Optical imaging; Testing;
Journal_Title :
Image Processing, IEEE Transactions on