基于改进R-L算法的运动模糊图像复原方法研究  

Research of Motion Blur Image Restoration Method Based on Improved R-L Algorithm

作　　者：陈员义 杨文福 周祥明 徐华银 YANG Yuanyi;YANG Wenfu;ZHOU Xiangming;XU Huayin(College of Internet of Things,Jiangxi Teachers College,Yingtan 335000,China;College of Aviation and Tourism,Jiangxi Teachers College,Yingtan 335000,China;Office of Teaching Affairs Jiangxi Normal University,Nanchang 330000,China)

机构地区：[1]江西师范高等专科学校物联网学院,江西鹰潭335000 [2]江西师范高等专科学校航空旅游学院,江西鹰潭335000 [3]江西师范大学教务处,南昌330000

出　　处：《兵器装备工程学报》2020年第10期228-232,共5页Journal of Ordnance Equipment Engineering

基　　金：江西省教育厅科学技术研究项目(GJJ151354)。

摘　　要：通过对运动模糊图像的复原过程进行分析,提出了一种基于二次变换改进的路西理查德森(Richardson-Lucy,R-L)算法图像复原方法,利用二次变换法精确估计出点扩散函数的系数,结合R-L算法完成退化图像的复原过程。通过仿真实验,量化对比了上述方法与传统R-L算法的复原效果,所述方法复原图像的灰度平均梯度以及峰值信噪比均提高约30%,实验结果表明,有效提升了运动模糊图像的复原精度,且针对含噪图像具有更高的鲁棒性。Based on the image point spread function(PSF)estimation and restoration algorithm,by analyzing the restoration process of motion blurred images,a method based on two-time transform parameter estimation and Richardson-Lucy(R-L)algorithm for image restoration were presented.Through simulation experiments,the advantages and disadvantages of the above method and the traditional image restoration methods were compared and analyzed,both the gray scale mean gradient and the Peak signal-to-noise ratio were increased by about 30%.The experimental results shown that the method can effectively improve the restoration accuracy of degraded images,and has higher robustness for noisy images.

关 键 词：运动模糊 点扩散函数 Richardson-Lucy算法 图像复原 二次变换法 

分 类 号：TJ306[兵器科学与技术—火炮、自动武器与弹药工程] TP391[自动化与计算机技术—计算机应用技术]