一维信号复原的几种迭代正则化方法比较研究

Several Iterative Regularization Methods for One-dimensional Signal Recovery

一维模糊含噪信号的复原是一类典型的不适定问题。经典的信号处理方法由于其自身局限性而制约了模糊含噪信号的复原效果。在分析模糊含噪信号特点的基础上,引入了3种不同的迭代正则化方法,并在不同噪声水平下比较了这3种正则化方法的信号复原效果。

The recovery of one-dimensional blurring-noise signal is a typical ill-posed problem.But the classic signal processing methods(as Wiener filter,Fourier transform and short-time Fourier transform et.al.)can't solve the problem effectively because of their limitations.Then three iterative regularization methods are introduced in this article based on analyzing the characters of the blurring-noise signals.While the recovery the three iterative regularization methods are compared by the experimental results under different noise levels.