Blur length estimation in linear motion blurred images using evolutionary algorithms

Abstract

Motion blur is a common blur type that created due to motion of camera or some objects in scene. The blur kernel, i.e. the function that simulate the motion blur process, depends on the length of blur. Therefore, the estimation of blur length is a vital and sensitive stage in the process of reconstructing a sharp version of a motion blurred image, i.e. image deblurring. In this paper, a method is proposed for estimation of the motion blur length using the evolutionary methods. To do this, we take the advantage of the relation between a blur metric, called NIDCT, and the blur length. Then this relation is learned via the evolutionary algorithms. The learned relation can be used to estimate the motion blur length in a blurred image. The efficiency of the proposed method is demonstrated by performing some quantitative and qualitative experiments.