Modeling Of Degraded Images By Complex Ambient Light Scattering And Its Restoration Method

The images restoration problem is an important research topic in the field of images processing research.This thesis focuses on a class of degraded images problems caused by ambient light scattering.In fact,atmospheric or underwater suspended solids in the natural environment often cause severe light scattering problems in natural conditions,such as haze weather images,sandstorm weather images and underwater images.In the digital imaging sampling process,the final images will have serious degradation problems,and the effectiveness of images will be seriously affected.In our thesis,we reconstruct a new imaging model related to ambient light scattering combining a large amount of imaging data.we analyze natural images,hazy images,and other degraded images subjected to severe ambient light scattering.We put forward a completely new prior called light intensity channel prior.Based on this prior,we present a imaging restoration method degraded by complex ambient light scattering.To illustrate the effectiveness and applicability of the results of this thesis,the numerical experiments in this paper are specifically divided into three parts: application testing of sandstorm images,and application testing of underwater images,application testing of other images degraded by ambient light scattering.The experimental results show that the new method can more effectively restore the images degraded by ambient light scattering,compared with some current restoration methods related to images degraded by ambient light scattering.