摘要
介绍了一种基于非线性多小波变换的信号去噪方法。通常子波域去噪中使用的Donoho软阈值法是很有效的,但是,由于忽略了边缘检测,导致在重构信号时丢失了部分的边缘信息,着眼于上述不足,在非线性多小波变换的基础上,提出了一种边缘检测与软阈值去噪相结合的去噪算法,实验证明此算法对提高重构信号的信噪比是非常有效的。
In this paper, a nonlinear multiwavelet transform based signal denoising method is introduced. Donoho抯 soft thresholding method used in wavelet denosing is effective. But it loses part of edge information in reconstruction due to neglecting singularity detecting. Because of this, we propose a new algorithm that is based on nonlinear multiwavelet transform and combines singularity detecting with soft thresholding. Examples show that the new algorithm is effective.
出处
《系统仿真学报》
CAS
CSCD
2003年第8期1074-1075,共2页
Journal of System Simulation
基金
国家自然科学基金资助项目(60172022)
关键词
软阈值
信号去噪
非线性多小波变换
边缘检测
soft thresholding
signal denoising
nonlinear multiwavelet transform
edge detecting
