测井信号的模极大值小波去噪与交替投影重构

Modulus Maxima Wavelet Denoising and Alternating Projection Reconstruction of Logging Signals

针对核磁共振测井自旋回波信号的特点,进行了模极大值小波变换,经过三层分解后,近似系数既能反映出原始信号的特征信息,且细节系数分布紧密,表明信号噪声较集中。通过交叉投影重构后的信号光滑、平稳,很好地体现了去噪后的信号与原始信号的相近性。

Considering the characteristics of nuclear magnetic resonance( NMR) logging's spin echo signals,the modulus maxima wavelet transform was implemented. After three layers of decomposition,the approximate coefficient can reflect the characteristic information of original signals; and the closely-distributed detail coefficients indicates a higher concentration of signal noise; and the smooth and steady signals under alternating projection and reconstruction can incarnate the proximity between the denoising signal and the original signal.