Error-parameter analysis based blind identification and super-resolution restoration algorithm

YAN Hua1; LIU Ju2; SUN Jian-de2; HUANG Qing-mimg3

您当前的位置：

首页 >

文章列表页 >

Error-parameter analysis based blind identification and super-resolution restoration algorithm

更新时间：2024-10-14

- Error-parameter analysis based blind identification and super-resolution restoration algorithm
- Vol. 30, Issue 8, Pages: 62-68(2009)
- 作者机构：
  
  1. 山东经济学院计算机科学与技术学院
  2. 山东大学信息科学与工程学院
  3. 西安电子科技大学智能信息处理研究所
- 作者简介：
- 基金信息：
- DOI：
  CLC： TP391.41
- Published：2009
- 稿件说明：
移动端阅览
YAN Hua1, LIU Ju2, SUN Jian-de2, et al. Error-parameter analysis based blind identification and super-resolution restoration algorithm[J]. 2009, 30(8): 62-68.
DOI：

YAN Hua1, LIU Ju2, SUN Jian-de2, et al. Error-parameter analysis based blind identification and super-resolution restoration algorithm[J]. 2009, 30(8): 62-68. DOI：

摘要

提出了一种实现规整化超分辨率复原算法的新方案

用来去除欠采样过程中的空间积分模糊和补零插值导致的边缘波纹。并将这种新方案运用到误差-参数分析法来准确地估计表征模糊函数的参数。当运动估计存在误差时

将误差-参数分析法和搜索算法相结合

降低参数辨识的计算量。仿真实验证明了提出的超分辨率算法实现方案和盲辨识方案的有效性。

Abstract

A new implementation scheme of regularization super-resolution restoration algorithm was addressed

which could remove spatial integration performed by the sensors in down-sampling process and restrain edge artifact caused by zero-padding up-sampling operation. The implementation scheme could be applied to error-parameter analysis for accurate blur identification. If motion estimation was inaccurate

searching algorithm could be combined with error-parameter analysis to reduce computation cost. Stimulation results demonstrate the effectiveness of the new implementation scheme and blur identification.

关键词

Keywords

references

Views

817

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Image restoration algorithm based on compensated transmission and adaptive haze concentration coefficient

Related Author

Yan YANG

Zhiwei WANG

Related Institution

School of Electronic and Information Engineering, Lanzhou Jiaotong University

AI问答

⁰