Image prior combination in space-variant blur deconvolution for the dual exposure problem

In this paper we propose a space-variant blur estimation and effective deconvolution method when combining a long exposure blurry image with a short exposure noisy one. The blur in the long exposure shot is mainly caused by camera shake or object motion, and the noise of the underexposed image is introduced by the gain factor applied to the sensor when the ISO is set to a high value. The image pair is divided in overlapping patches for processing. The main idea in this work is to incorporate a combination of prior image models to a spatially-varying deblurring/denoising framework which is applied to each patch. The method exploits kernel and parameters estimation to choose between denoise or deblur each patch. In addition, the proposed approach estimates all necessary parameters automatically without user supervision. Experiments on both synthetic and real images validate the used approach.