摘要
本文提出了基于小波标架的软门限压缩降噪方法，它是根据小波标架表示系数在不同的尺度有不同的方差，在各尺度分别模拟理想选择估计器来选择门限对小波系数进行压缩。该方法能有效地减少［３，４］中基于小波基方法的振荡效应，并对仿真结果作了理论分析。
In this paper, we prsent a signal de-noising method by soft-threshold shrinking based on descrete wavelet frame transform. The threshold is selected by imitating ideal selective estmator, and wavelet coefficients are shrunk by soft-threshold shrinkage functions which depend on threshold and noise variance in wavelet transform domain at each scale. This mehtod can reduce the oscillating effects which are exhibited in the vicinity of signal discontinuities, and we give the theoretic interpretation of the experimental results.
出处
《信号处理》
CSCD
1999年第3期204-211,共8页
Journal of Signal Processing
关键词
离散小波标架
降噪
信号处理
滤波
Discrete wavelet frame, De-noising,Soft-threshold shrinking
