Estimating an Image's Blur Kernel from Edge Intensity Profiles.

摘要

This paper presents a simple and fast method to estimate the blur kernel model, support size, and its parameters directly from a blurry image rather than relying on the standard models. Also, this method estimates the parameters without the need to search the parameter space. In addition, this edge profile method is local and can provide a measure of spatial variation. The main insight of this work is that the profile of a blurry edge in an image is equivalent to a cumulative distribution function, which is used to estimate the underlying blur kernel functional form and parameters. We show how to utilize the concepts behind the statistical tools for fitting data distributions to analytically obtain an estimate of the blur kernel that incorporates blur from all sources, including factors inherent in the imaging system. The validity of this method is demonstrated with idealized and standard images and then NIR, SWIR, MWIR, and Active IR imagery results are shown to be similar to results from state-of-the-art methods.