ABSTRACT

In the previous chapter, we considered the problem of super-resolution reconstruction of images with a priori information. In the general solution of the super-reolution reconstruction problem, we deal with three degradation phenomena: general geometric registration warp, blurring, and additive noise. We assumed that these degradation phenomena were known a priori. There are applications in which the degradation phenomena are unknown and difficult to estimate within the super-resolution reconstruction process. Obtaining a high-resolution (HR) image from degraded low-resolution (LR) images with unknown degradation phenomena is known as blind super-resolution reconstruction of images. Few methods deal only with blind super-resolution reconstruction [114–123].