Abstract

In this paper, we consider the nonnegatively constrained multichannel image deblurring problem and propose regularizing active set methods for numerical restoration. For image deblurring problems, it is reasonable to solve a regularizing model with nonnegativity constraints because of the physical meaning of the image. We consider a general regularizing $l_p-l_q$ model with nonnegativity constraints. For p and q equaling 2, the model is in a convex quadratic form, therefore, the active set method is proposed since the nonnegativity constraints are imposed naturally. For p and q not equaling 2, we present an active set method with a feasible Newton-conjugate gradient solution technique. Numerical experiments are presented for ill-posed three-channel blurred image restoration problems.

© 2009 Optical Society of America

A limited-memory, quasi-Newton preconditioner for nonnegatively constrained image reconstruction

Johnathan M. Bardsley
J. Opt. Soc. Am. A 21(5) 724-731 (2004)

Selection of regularization parameter in total variation image restoration

Haiyong Liao, Fang Li, and Michael K. Ng
J. Opt. Soc. Am. A 26(11) 2311-2320 (2009)

Regularizing active set method for retrieval of the atmospheric aerosol particle size distribution function

Yanfei Wang and Changchun Yang
J. Opt. Soc. Am. A 25(2) 348-356 (2008)

Hybrid regularization method for the ill-posed inversion of multiwavelength lidar data in the retrieval of aerosol size distributions

Christine Böckmann
Appl. Opt. 40(9) 1329-1342 (2001)

Sparsity constrained regularization for multiframe image restoration

Premchandra M. Shankar and Mark A. Neifeld
J. Opt. Soc. Am. A 25(5) 1199-1214 (2008)