Restoration of Atmospherically Degraded Images using aSparse Prior

摘要

The reconstruction of turbulence-affected images has been an active research topic in the field of astro-nomical imaging. Many approaches have been proposed in the literature. Recently, researchers haveextended the methods to the recovery of long-path territorial natural scene surveillance, which is af-fected even more by air turbulence. Some approaches from astronomical imaging also work well in thelatter problem. However, although these methods have involved statistics, such as a statistical model of atmosphericturbulence or the probability distribution of photons forming an image, they have not taken account ofthe statistical properties of natural scenes observed in long-path horizontal imagery. Recent researchby others has made use of the fact that a real world image generally has a sparse distribution ofits derivatives. In this paper, we investigate algorithms with such a constraint imposed during therestoration of turbulence-affected images. This paper proposes an iterative, blind deconvolution algorithm that follows a registration andaveraging method to remove anisoplanatic warping in a time sequence of degraded images. The useof a sparse prior helps to reduce noise, produce sharper edges and remove unwanted artifacts in theestimated image for the reason that it pushes only a small number of pixels to have non-zero (or large)derivatives. We test the new algorithm with simulated and natural data and experiments show that itperforms well.