图像盲复原问题的研究

Research blind image restoration problem

大气湍流、相对运动、散焦和噪声等诸多因素会导致获取的图像产生降质。图像盲复原的任务是在预定义的意义上改善退化模型未知的图像。本文深入分析图像的退化模型、噪声模型及复原算法后,基于迭代盲解卷积算法,对含高斯噪声的且由大气湍流或悬浮物散射造成的降质图像提出了一种复合的盲复原方法。同时分析了噪声抑制对改善病态性问题的影响。实验结果表明:该方法复原效果较迭代盲解卷积算法、MATLAB库函数deconvblind()方法更佳。

The presence of the atmospheric interference, relative motion defocusing,noise and many other factors led to decline in images. The ultimate goal of restoration techniques is to improve an image in some predefined sense. But for the cause of some of the parameters of degradation is not clear. In this paper, the model of degradation and noise, and common restoration methods and evaluation was studied. Based on the research on blind deconvoiution using for image, This paper presents a composite restoration method. Through the simulation test noise suppression technology to improve the morbidly was analyzed, and it is better than IBD algorithm or MATLAB library function method.