超分辨率图像重建方法及其在超声图像中的应用  

Super-resolution Image Reconstruction and Its Application in Ultrasound Images

作　　者：孙志民[1] 汪源源[1] 王威琪[1] 

机构地区：[1]复旦大学电子工程系,上海200433

出　　处：《仪器仪表学报》2004年第z3期363-365,371,共4页Chinese Journal of Scientific Instrument

基　　金：上海市曙光计划资助项目。

摘　　要：采用时域模型实现基于多帧图像的超分辨率图像重建方法。方法的步骤为:先对多帧图像进行光流场配准,用基于最速下降的最大似然估计法重建超分辨率的图像,并对重建的图像进行基于维纳滤波的降噪。在模拟实验中,给定一幅高分辨率源图像,模拟成像过程获取具有亚像素平动的低分辨率加噪图像序列,然后利用上述方法重建超分辨率图像。对标准数字图像进行模拟实验,结果表明该重建方法具有良好的效果。对超声图像的处理结果也表明:该方法能有效提高超声图像的分辨率,有望为超声图像分辨率增强提供一种新的手段。A multi-frame based super-resolution image reconstruction method is described under a time-domain model.A sequence of low-resolution images was firstly registered with the optic flow method.Then the maximum likelihood (ML) estimation based on the steepest descent (SD) algorithm was applied to reconstruct the super-resolution image.The reconstructed result was finally de-noised by a Wiener filter.In the experiments,a certain high-resolution image was given as the source image of the simulated imaging process,from which a low-resolution noisy image sequence with sub-pixel translation was obtained.The super-resolution image was reconstructed by this described method.The performance of this method in the experiments on several standard digital images was excellent.The reconstruction results from the simulated ultrasound image sequence also show that this method can effectively improve the resolution of ultrasound images and can be expected as a novel method for the resolution enhancement of ultrasound images.

关 键 词：超分辨率 时域模型 维纳滤波 超声图像 

分 类 号：TH7-55[机械工程—仪器科学与技术]