Blind infrared images reconstruction using covariogram regularization from a regular pentagon

Abstract

Infrared images reconstruction is a fundamental problem in computer vision, which can remove the random image noise and produce sharp features. In this study, we propose a novel infrared image reconstruction model with covariogram regularization. Firstly, the covariogram of a regular pentagon is constructed. To enlarge the class of domains for which the forms of covariogram can be written explicitly, we reveal the explicit form of the covariogram for a regular pentagon in terms of the restricted chord function. As an application of the covariogram, the explicit expression of the containment function is achieved for a regular pentagon. Then, due to the non-convex property of the covariogram regularization, the fused coordinate descent framework is introduced to optimize the proposed image reconstruction model. To demonstrate the good performance of the proposed method, several real infrared image experiments are carried out. It shows that proposed method can smooth the random noise effectively and preserve the image structures.