计算机工程与应用 ›› 2010, Vol. 46 ›› Issue (9): 146-148.DOI: 10.3778/j.issn.1002-8331.2010.09.041

• 图形、图像、模式识别 • 上一篇    下一篇

一种提高SVD滤波性能的新方法

杨敦旭1,杜晓凤2,李翠华1   

  1. 1.厦门大学 计算机科学系,福建 厦门 361005
    2.厦门大学 自动化系,福建 厦门 361005
  • 收稿日期:2009-07-14 修回日期:2010-01-13 出版日期:2010-03-21 发布日期:2010-03-21
  • 通讯作者: 杨敦旭

Novel approach to SVD-based image filtering improvement

YANG Dun-xu1,DU Xiao-feng2,LI Cui-hua1   

  1. 1.Department of Computer Science,Xiamen University,Xiamen,Fujian 361005,China
    2.Department of Automation,Xiamen University,Xiamen,Fujian 361005,China
  • Received:2009-07-14 Revised:2010-01-13 Online:2010-03-21 Published:2010-03-21
  • Contact: YANG Dun-xu

摘要: 提出一种提高SVD滤波性能的新方法。基于奇异值分解滤波可以有效地分析水平(垂直)方向的图像特性。根据图像的局部方向,自适应地调整待滤波区域的形状,使重采样后局部区域中的边缘垂直或水平,再对局部区域进行奇异值分解滤波;所得的结果加权平均,得到信号估计值。将这一算法应用于图像去噪,实验结果表明,新方法可以有效地提高SVD的滤波性能。

关键词: 奇异值分解, 图像去噪, 方向检测, 自适应滤波

Abstract: A novel approach to improve the filtering efficiency of a noisy image is proposed.Image filtering based on SVD favors the denoising in the line(horizontal) and column(vertical) direction.Based on this property,the new denoising method adapts shape and size of block to local orientation before performing SVD filtering.Through over-complete representation in overlap regions,the proposed method performs well in denoising and preserving image details.

Key words: Singular Value Decomposition(SVD), image denoising, orientation detection, adaptive filtering

中图分类号: