摘要
运用小波框架的复原法恢复运动模糊图像,在紧小波框架系统下,用正则化方法对模糊图像进行处理,将盲解卷积转化为极小问题,利用分裂Bregman迭代法求解去模糊图像。对直线匀速运动、曲线匀速运动和任意曲线运动模糊图像进行实验,取得了良好效果。
Motion image deblurring is one of fundamental problem in digital image processing. In this paper, a novel de- blurring method is proposed which is based on the wavelet framework. Under tight wavelet frame systems, a reglization of sparsity is imposed on both the blurred image and the motion-blurred kernel. Blind image deblurring was transformed as a minimum problem and the split Bregman was applied to solve this problem. Experiments prove our method is very effec tive on image deblurring even thought there is no any priori on the motion-blurred kernel.
出处
《软件导刊》
2016年第7期186-189,共4页
Software Guide
关键词
图像复原
盲卷积
正则化方法
Image Deblurring
Blind Deconvolution
Regularization Method
