基于OpenCV的匀速模糊图像复原与算法分析  

Restoration and Algorithm Analysis of Uniform Motion-Blurred Image Based on OpenCV

作　　者：程凡永[1] 罗海波[1] 

机构地区：[1]闽江学院计算机科学系,福建福州350108

出　　处：《湖南工业大学学报》2011年第3期59-63,共5页Journal of Hunan University of Technology

基　　金：闽江学院育苗基金资助项目(YKY1004);福建省教育厅教学改革基金资助项目(闽教高[2007]4号)

摘　　要：对水平匀速直线运动造成的模糊图像进行频谱分析,估计出模糊距离,进而求得图像的点扩散函数。采用维纳滤波法和投影复原法复原图像,分析这2种算法对模糊距离的鲁棒性。采用OpenCV与MFC函数库作为软件设计基础,实现的软件系统对匀速模糊图像的复原效果比较理想。Through analyzing the spectrum of blurred image caused by horizontal uniform linear motion,the blurred distance can be estimated,and furthermore the point spread function of the image could be obtained.The wiener filtering method and projection restoration method are applied to restore the image and the robustness of these two approaches is analyzed.Adopting OpenCV and MFC function library as the basis for software design,the implemented software system achieves ideal image restoration of motion-blurred image.

关 键 词：OPENCV 点扩散函数估计 维纳滤波法 投影复原法 鲁棒性 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]