| 摘要: |
| 采用直接复原方法和插值复原方法对单幅欠采样低分辨率模糊图像进行复原实验,以考察线性插值在图像复原中的作用。直接复原法是对采集到的欠采样低分辨率模糊图像直接用最大似然图像复原算法进行复原,不作任何预处理。插值复原法是先对图像进行插值预处理,增加像素数,提高图像采样分辨率,然后再用最大似然图像复原算法进行图像复原。结果是直接复原方法和插值复原方法对欠采样模糊图像的复原,效果相同,主观视觉和评价数据十分接近。这说明通常用于单画幅欠采样模糊图像复原预处理的线性插值没有起到提高复原效果的作用。 |
| 关键词: 插值 欠采样图像 图像复原 |
| DOI： |
| 投稿时间：2006-07-17 |
| 基金项目: |
|
| Effect of Interpolation on Restoration of Single-frame Undersampled Image |
|
SU Hong-wei, CHEN Hua
|
| (School of Computer, Electronics and Information, Guangxi University, Nanning, Guangxi, 530004, China) |
| Abstract: |
| The direct-restoration method and the interpolation-restoration method are adopted in the experiment to deal with single-frame undersampled blurry image of low resolution in order to review the effect of linearity interpolation on image restoration.In the direct-restoration,the undersampled blurry image of low resolution is restored directly by maximum-likelihood algorithm(ML) without any preprocessing.In the interpolation-restoration,the undersampled blurry image of low resolution is preprocessed by the interpolation to increase image pixels and improve image sampling resolution before it is restored by ML algorithm.The results show that the subjective vision and external estimation index of the image in these two methods are similar.It is revealed that the linearity interpolation can not improve image restoration on single-frame undersampled blurry image. |
| Key words: linearity interpolation undersampled image image restoration |
