J4 ›› 2012, Vol. 39 ›› Issue (6): 154-161.doi: 10.3969/j.issn.1001-2400.2012.06.025

• 研究论文 • 上一篇    下一篇

经验模式分解域自适应滤波方法

李烨;杨家玮;安金坤;梁彦霞   

  1. (西安电子科技大学 综合业务网理论及关键技术国家重点实验室,陕西 西安  710071)
  • 收稿日期:2011-05-10 出版日期:2012-12-20 发布日期:2013-01-17
  • 通讯作者: 李烨
  • 作者简介:李烨(1982-),男,博士,E-mail: wit951@sohu.com.
  • 基金资助:

    国家自然科学基金资助项目(61072068);国家杰出青年科学基金资助项目(60725105);长江学者和创新团队发展计划资助项目(IRT0852)

Adaptive filtering method in the empirical mode decomposition domain

LI Ye;YANG Jiawei;AN Jinkun;LIANG Yanxia   

  1. (State Key Lab. of Integrated Service Networks, Xidian Univ., Xi'an  710071, China)
  • Received:2011-05-10 Online:2012-12-20 Published:2013-01-17
  • Contact: LI Ye

摘要:

为了利用经验模式分解法提取信号边缘信息,提出一种基于经验模式分解的自适应滤波方法,并给出了噪声功率阈值的两种选取方法.该滤波方法首先对信号进行经验模式分解; 其次对相邻尺度上残差分量一阶导数信号进行空间相关性计算,并对归一化空间相关函数与残差分量一阶导数进行逐点比较,实现对残差分量一阶导数的滤波; 最后根据噪声功率阈值判断自适应滤波过程是否结束.仿真实验结果显示,本方法可以准确提取信号边缘信息,同时抑制噪声信号.

关键词: 边缘检测, 多尺度, 空间相关性, 自适应滤波, 门限, 经验模式分解

Abstract:

An adaptive filtering method in the empirical mode decomposition domain is proposed to detect the edges of a noisy signal, and two methods for selecting the noise power threshold are presented. Firstly, the adaptive filtering method decomposes the noisy signal by empirical mode decomposition; secondly, the spatial correlation of the first derivatives of residuals in adjacent scales is calculated; thirdly, the normalized spatial correlation is compared with the first derivative point by point to achieve the filtering; Finally, the noise power threshold decides whether the filtering should be finished or not. Simulation tests show that the new adaptive filtering method can accurately detect edges of signals, and suppress the noise.

Key words: edge detection, multi-scale, spatial correlation, adaptive filtering, threshold, empirical mode decomposition

中图分类号: 

  • TN911.7