Study On Realization And Denoising Of Ultrasonic Phased Array Total Focusing Method Imaging

Ultrasonic phased array is an important testing method in the filed of industrial nondestructive testing.By using the Total Focusing Method in ultrasonic phased array system,the virtual focus of acoustic beam can be formed in each pixel of the imaging area,which can effectively improve the imaging quality and facilitate the resolution and detection of defects.This paper in domestic and foreign research present situation investigation and understand the Total Focusing Method,and analyzing the source and characteristics of noise,this paper design a TFM imaging and denosing scheme based on FPGA platform,and verifies the imaging effect and signal-to-noise ratio based on 16×64 ultrasonic phased array system.In view of the characteristics of the Total Focusing Method suitable for parallel computing and pipeline processing,A scheme of composite image calculation based on FPGA is designed.The idea of modular design is adopted to divide the TFM into A-scan data synthesis module,de-baselines module and envelope detection module.A sweep data synthesis module is composed of sound path calculation and stacking synthesis.By calculating sound path,it is converted into address index of RAM to perform stacking synthesis on the whole matrix data set.The de-baselining module adopts the moving mean de-baselining algorithm to perform preliminary filtering and de-baselining of the original TFM image obtained by superposition.The envelope detection module realizes the Hilbert variable exchange envelope algorithm based on FFT,which effectively improves the resolution of defect information.According to the noise source and characteristics of TFM imaging,a denoising scheme based on wavelet threshold denoising method is designed.Through simulation and comparison,the influence of different wavelet bases and threshold algorithms on imaging quality is verified,and the realization scheme of LWT threshold denoising based on FPGA platform is designed.The denoising of TFM image is divided into threshold calculation processing module,LWT decomposition module and LWT reconstruction module.The threshold calculation processing module obtains the median approximation by statistical method,and then obtains the threshold estimation,and denoises the wavelet high frequency coefficient.The LWT decomposition and reconstruction module adopts pipelined architecture to carry out 4-level decomposition,noise reduction and reconstruction of the signal,so as to obtain the TFM imaging image after noise reduction.Based on the 16×64 ultrasonic phased array system,this paper verifies the imaging effect and noise reduction effect of the TFM image.The test results show that the TFM module can reflect the defect information well,and the SNR of the image is improved to more than 6 d B after denoising by the LWT threshold denoising module.