小波变换模极大值消噪算法的研究

根据信号和噪声在小波变换下的不同特性,采用模极大值去噪的算法.其中包括Mallat交替投影算法和模极大值小波域去噪算法,Mallat交替投影算法是通过剔除那些幅度随尺度而减小的模极大值点,保留幅度尺度而增大的模极大值点,然后将剩余的模极大值利用模极大值重构算法来重建原始信号,达到滤波去噪的目的.而模极大值小波域算法是先用Adhoc算法求出信号的模极大值,再根据模极大值小波域的定义求出信号的模极大值小波域,从而得到信号的小波系数,然后逆变换得到信号.实验结果表明两种算法都具有很好的消噪效果.

Denoising Algorithm Based on Wavelet Transform Modulus Maximum

Based on the different properties of signals and noise with wavelet transform, the wavelet transform modulus maximum is presented to eliminate the noise. It includes Mallat alternative projection algorithm and denoising algorithm based on modulus maximum in wavelet field. The first method is put forward to remove the modulus maximum that decreases with increasing scale and reserve those to the contrary, and then, the remainder is reconstructed to the original signal so as to denoise. The wavelet field modulus maximum is realized by obtaining modulus maximum according to Adhoc algorithm. Its definition is computed to get the wavelet coefficient. The singularity is reconstructed through inverse wavelet transform. Experiment result shows that two methods can suppress the noise effectively.