运动模糊图像复原算法的改进及性能研究  

Improvement and Performance Research of Motion Blurred Image Restoration Algorithm

作　　者：刘红亮[1] 陈维义[1] 傅茂聪 

机构地区：[1]海军工程大学兵器工程系,湖北武汉430033

出　　处：《电视技术》2015年第16期43-45,50,共4页Video Engineering

基　　金：装备预研基金资助项目(9140A09031213JB11001)

摘　　要：为提高运动模糊图像复原算法的有效性和实时性,分析了常见图像复原算法的优缺点。在此基础上,提出了一种基于小波分解和维纳滤波相结合的图像复原算法。新算法充分利用了小波变换的多分辨率分析特性和维纳滤波复原算法的高效性,既有效抑制了噪声,又减小了图像的灰度失真。仿真结果表明,该算法不仅提升了图像复原质量,也能满足系统实时性的要求,是一种有效的图像复原算法。In order to improve the effectiveness and real-time performance of motion blurred image restoration algorithm, the advan- tages and disadvantages of common image restoration algorithm are discussed in this paper. After that, a new image restoration al- gorithm based on Wiener filtering and wavelet decomposition is proposed. The multi-resolution analysis of wavelet transform and the efficiency of Wiener filtering are made full use of by the new algorithm. Therefore, the noise and grayscale distortion of tile mo- tion blurred image is restrained. Simulation results show that the proposed algorithm is an efficient image restoration algorithm, which promotes the quality of image restoration and meets the requirements of real-time system.

关 键 词：运动模糊图像 维纳滤波 小波分解 实时性 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]