基于小波的多路信号估计方法

从带噪信号中估计源信号在很多应用领域都是重要课题,例如通信、医学成像、远程图像以及天文学等等。估计多路信号是其中的一个小子集。本文阐述了最近提出的基于小波的多路二维信号估计方法。这种方法通过离散傅立叶变换(DFT)和二维离散小波变换(DWT)近似最佳地使数据解相关,从而充分地利用了信道内和信道间信号的相关性。同时,运用Normal Shrink降噪算法改进估计结果。

Wavelet-based Method for Multichannel Signal Estimation

Estimation of signals from noisy observations is an important topic with applications in areas such as communication, medical imaging,remote imaging,and astronomy.A subset of this broad problem is the estimation of multichannel signals.In this paper, we illustrate a recently proposed wavelet-based method for estimating signals from noisy 2-D multichannel observations.This method ex- ploits both the intrachannel and interchannel signal correlations using a discrete Fourier transform(DFF)and 2-D discrete wavelet trans- form(DWT)to approximate the optimal data decorrelation.The proposed method called NormalShrink is used to improve the estimate re- sults.