红外运动模糊图像复原技术  

Restoration of Motion-blurred Infrared Image

作　　者：杨晓冬[1] 刘俊[1] 张倩倩[1] 

机构地区：[1]杭州电子科技大学信息与控制研究所,浙江杭州310018

出　　处：《杭州电子科技大学学报（自然科学版）》2012年第4期144-147,共4页Journal of Hangzhou Dianzi University：Natural Sciences

基　　金：总装备部"十二五"装备预研资助项目(*******)

摘　　要：该文分析了红外图像运动模糊产生的原理,建立运动模糊及复原数学模型。复原过程分包括构建点扩散函数和维纳滤波两部分。采用方向微分,双线性插值法求取模糊方向,通过微分自相关函数估算模糊尺度,从而构建最为近似的二维匀速直线运动的点扩散函数,并对运动模糊图像进行维纳滤波的复原。实验结果表明,针对有噪声的图像运动模糊复原,维纳滤波复原图像效果好,抗噪能力强。In this paper, the mechanism of a motion-blurred image generated from infrared imaging system was introduced. The mathematical model of direction differential, bilinear interpolation method and the related function to obtain the point spread function of two-dimensional uniform linear motion. It also explicates Wiener filter for motion-blurred infrared images, and compares their advantages and disadvantages. Finally, the algo- rithm is demonstrated through a series of experiments which reveal that the Wiener filter for restoration motion- blur imaging is more better than inverse filtering.

关 键 词：图像复原 运动模糊 点扩散函数 方向微分 维纳滤波 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]