Abstract: Aiming at effectively filtering the noise components of pulsed eddy current testing signal in ferromagnetic materials, the method of wavelet analysis and Kalman filter denoising is proposed. The basic principle of wavelet threshold denoising is expounded and the threshold parameters of wavelet denoising are optimized through the signal denoising experiment. So the optimal wavelet threshold denoising is implemented. Because there exists still a small amount of non-stationary noise signal after wavelet denoising, the method of Kalman filter denoising is proposed. The signal denoising experimental results show that Kalman filter can effectively filter the noise components, increase the signal to noise ratio (SNR) to about 34.976 7 dB and reduce the root mean square error (RMSE) to about 0.0493 95.