基于Split Bregman算法的图像复原

图像复原研究由来已久。一种比较传统的图像复原方法是假设系统的脉冲响应,即事先知道点扩散函数。然而,在大多数实际情况下,很难具体确定点扩散函数,我们仅能通过退化之后的图像来恢复图像,即图像盲复原法。图像盲复原法是在已经观测到的退化图像的基础上,根据特定条件建立合适的数学模型,从而在一定条件下获得最接近原始图像的逼近值。文章通过Split Bregman算法对灰度图和LAB图进行图像复原,并进行了对比实验。

Image restoration based on Split Bregman algorithm

The research of image restoration began very early.Many years ago,the image restoration method assumed the impulse response of the system,that is,the point spread function was known in advance,which is a more traditional image restoration method.However,in most practical cases,it is difficult to determine the point spread function.We can only restore the image through the degraded image.This method is called blind image restoration.Blind image restoration method is to add specific conditions on the basis of the observed degraded image,so as to establish an appropriate mathematical model and obtain the approximation value closest to the original image under certain conditions.In this paper,split Bregman algorithm is used to restore gray image and lab image,and comparative experiments are carried out.