计算机工程与应用 ›› 2014, Vol. 50 ›› Issue (11): 223-228.

• 信号处理 • 上一篇    下一篇

一种有效的脉冲噪声去除算法

胡积宝,朱  柱   

  1. 安庆师范学院 物理与电气工程学院,安徽 安庆 246011
  • 出版日期:2014-06-01 发布日期:2015-04-08

Research of impulse noise removal algorithm

HU Jibao, ZHU Zhu   

  1. School of Physics and Electronic Engineering, Anqing Normal University, Anqing, Anhui 246011, China
  • Online:2014-06-01 Published:2015-04-08

摘要: 噪声像素的检测精度是修复图像的重要因素。为提高噪声像素的检测精度,提出一种有效的两步脉冲噪声滤波算法,算法先以空间局部异常测度为基础,结合S-估计统计量对图像中的脉冲噪声进行检测,然后通过建立、解算正则化目标函数实现对噪声像素的赋值修复。实验验证了该算法的有效性。

关键词: 脉冲噪声, 空间局部异常测度, S-估计, 正则化

Abstract: An impulse noise removal algorithm based on spatial local outlier measurement is proposed, which can realize the elimination of the impulse noise in image effectively. The algorithm identifies impulse-noise in a corrupted image based on SLOM, S-estimate firstly. Then it restores noise candidates by establishing and solving the regularization objective function afterward. Extensive experimental results show the algorithm’s effectiveness.

Key words: impulse noise, Spatial Local Outlier Measure(SLOM), S-estimate, regularization