基于增量维纳滤波和空间自适应规整的超分辨率图像复原  

Incremental Wiener Filter and Space-Adaptive Regularization Based Super-Resolution Image Restoration

作　　者：闫华[1] 刘琚[1] 李道真[1] 孙建德[1] 

机构地区：[1]山东大学信息科学与工程学院,济南250100

出　　处：《电子与信息学报》2005年第1期35-38,共4页Journal of Electronics & Information Technology

基　　金：山东省自然科学基金(Y2000G12)山东省优秀中青年科学家科研奖励基金(01BS04)资助课题

摘　　要：该文提出一种新的超分辨率图像复原方案,该方案利用了空间自适应规整方法的空间分段平滑特性和增量维纳滤波的快速收敛能力,将解的先验约束结合到迭代过程中,通过多幅低分辨率降晰图像来估计一幅高分辨率图像。计算机仿真结果表明,该方法可以有效实现超分辨率图像复原,有效地提高了复原的自适应控制能力和收敛性能。A new method of super-resolution image restoration is proposed in the paper. In this method, the piecewise smoothness of space-adaptive regularization combines with the rapid convergence of incremental Wiener filter and through an iteration process with a priori constraint about image, one super-resolution image is estimated from several low-resolution images. Simulations show that the method can implement restoration effectively and has good convergence.

关 键 词：超分辨率 图像复原 增量维纳滤波 空间自适应规整 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]