J4

• 计算机科学 • 上一篇    下一篇

基于L-曲线正则化的MAP超分辨率图像复原

王宏志, 赵 爽, 胡艳君   

  1. 长春工业大学 计算机科学与工程学院, 长春 130012
  • 收稿日期:2007-10-26 修回日期:1900-01-01 出版日期:2008-03-26 发布日期:2008-03-26
  • 通讯作者: 王宏志

Superresolution Image Restoration with MAP Based on L-Curve

WANG Hongzhi, ZHAO Shuang, HU Yanjun   

  1. Institute of Computer Science and Engineering, Changchun University of Technology, Changchun 130012, China
  • Received:2007-10-26 Revised:1900-01-01 Online:2008-03-26 Published:2008-03-26
  • Contact: WANG Hongzhi

摘要: 提出一种基于L曲线正则化参数选择方法的最大后验 概率超分辨率图像复原算法. 该方法将最大后验概率估计和L-曲线正则化参数选择方法相结合, 可以有效地减少和去除复原图像中的噪声, 提高图像复原质量, 并具有较好的超分辨率复原能力.

关键词: 超分辨率, 图像复原, 最大后验概率, L-曲线

Abstract: We proposed an Bayesian maximum Aposteriori (MAP) estimator with L-curve regularization for superresolution restoration which incorporated maximum Aposteriori and L-curve method. Experiments demonstrate that this method can effectively reduce and remove noise in the restoration of images, get good quality restoration images and has high super resolution performance.

Key words: superresolution, image restoration, maximum Apost eriori, L-curve

中图分类号: 

  • TP317.4