Deconvolution of Neutron Degraded Images: Comparative Study between TSVD, Tikhonov Regularization and Particle Swarm Optimization Algorithm

Image deconvolution is an important problem in image processing. It is an ill-posed inverse problem, so regularization techniques are used to solve this problem by adding constraints to the objective function. Various popular algorithms have been developed to solve such problem. This paper proposes a new approach to the nonlinear neutron degraded images restoration problem which is useful in many images enhancement applications, based on swarm intelligence. We use the particle swarm optimization (PSO) applied for total variation (TV) minimization, instead of the standard Tikhonov regularization method. In this work, we attempt to reconstruct or recover neutron radiography images that have been degraded during acquisition; using some a priori knowledge of the degradation phenomenon. The truncated singular value decomposition (TSVD) method is also considered for image deconvolution in this paper. A comparison between the five methods is conducted, using several images.