Research On Signal Denoising And Phase Correction Based On Ultrasonic Synthetic Focusing

Compared with conventional ultrasound imaging, Ultrasonic synthetic aperture focusing technique (SAFT) has a higher resolution, but the ultrasound propagation in the material will produce diffraction, scattering, absorption phenomenon, coupled with poor coupling, electrical noise, and other factors, so that the collected ultrasonic signal affected by noise and phase aberration and other adverse factors will result in lower resolution and signal to noise ratio of imaging. So noise and phase distortion in the ultrasound signal will be discussed in the following thesis.In this thesis, ultrasonic synthetic aperture focusing technique and four improved technologies were researched discussed and their beam-forming processes are all deduced and simulated. In terms of their beam forming angle, the advantages and disadvantages are analyzed. The research will be done with synthetic data collection focused manner.Noise suppression process based on wavelet transform is analyzed in detail. Random noise and Gaussian noise will be added in the simulated Field Ⅱ ultrasonic signals which will be used in the noise suppression process by using Daubechies wavelet function and Symlet wavelet function with different orders respectively. Then the obtained wavelet function with a better noise suppression performance will be applied to the collected actual synthesis focused signal. Now, the ultrasonic signal noise suppression performance of this two kinds of wavelets functions will be analyzed.Analysis of the reason causing the phase aberration will be done, and mathematical model is established. Based on signal correlation phase correction method, analysis of the situation of each array element phase aberration between the received signals has been done. Delay error for each channel will be obtained by calculating the position of the maximum correlation coefficient between the signals. Through reverse method signal phase aberration will be eliminated, then the signals was used to reconstructed ultrasonic image with synthetic focusing delay and superposition method.According to the above method, db12 wavelet function and sym16 wavelet function, with a 20.9276dB and 22.0382dB SNR respectively, have a better performance on random noise suppression. db13 wavelet function and sym14 wavelet function, with a 21.0427dB and 21.2014dB SNR respectively,have a better performance on Gaussian noise suppression. For actual data, Symlet wavelet function is more suitable for the actual ultrasonic signal noise suppression. Our signal to noise ratio can improve 18.1994dB by using Wavelet Noise Reduction and phase-corrected firstly and synthetic focus imaging secondly. Studies have shown that, the method studied in this thesis can improve the signal to noise ratio and be used in practical application as a reference.