Research On The Restoration Of Motion-blurred Images

Image restoration is the process of transforming the input degraded images to the latent ones.The main reason that causes the degradation is due to the inevitable camera shake and environmental noise.These factors can be represented by a blur kernel called PSF?point spread function?,and in this article we focus on the uniform motion blur restoration.This problem can be categorized into two sections,one is the non-blind deconvolution,which means PSF is a known variable,the goal of which is to estimate the latent image using the input blurred image and the blur kernel.The other section is the blind deconvolution,the goal is to estimate both the latent image and the blur kernel given only the input motion-blurred image.This paper researches the uniform motion-blurred single image blind deconvolution algorithm.For uniform system,it can be modeled by the convolution of two terms,however since it's an ill-posed problem,we construct the MAP framework in this paper to suppress the ringing effects,L₀-norm and L₁-norm are utilized for the regularization term of latent image and PSF separately.At last we conduct comparative experiments to illustrate that the deblurring result of the proposed method exceeds other deblurring algorithms.