Application of Wavelet Threshold De-noising Method in Electromagnetic Ultrasonic Signal Processing

摘要

In order to solve the problem of containing noise in echo signal detected by electromagnetic ultrasonic metal plate thickness, wavelet threshold method was used for de-noising. First to choose the wavelet function and decomposition scale to decompose the echo signal, then structure threshold rules by the weighted average method and dispose the wavelet coefficient obtained by decomposition, reconstruction signal. De-noising signal acquired by MATLAB simulation had obvious effect and SNR was 9.4441, RMSE was 0.2. Experimental results show that wavelet threshold value method can effectively inhibit noise doped echo signal and improve electromagnetic ultrasonic detection accuracy.