基于自适应小波阈值的超声信号消噪

在铝合金锻件的超声无损检测中为了消除晶粒散射引起的相干噪声，通过建立缺陷回波检测数学模型，提出了一种基于新阈值函数的Stein无偏风险估计自适应消噪方法.利用新的阈值函数得到离散小波变换各尺度下的小波系数，对小波阈值进行最小均方误差意义上的迭代，基于小波系数估计值进行离散小波反变换以得到信号的估计值，通过反复迭代运算得到缺陷回波的最优消噪模型.对含缺陷铝合金锻件的超声信号处理实验结果表明，与常用的固定硬、软阈值相比，自适应消噪方法能够更好地去除散射噪声及增强缺陷信回波信号.

Adaptive wavelet thresholding based ultrasonic signal denoising

A model of ultrasonic flaw echo detection was built to depress the high scattering microstructure noise in the ultrasonic signal from aluminum alloy forge.A new wavelet thresholding function based Stein unbiased risk estimation method was set up to perform the adaptive denoising procedure.The discrete wavelet coefficients of different scale were calculated by using the new thresholding function,then a thresholds iterative process based on the minimum mean square error was implemented.The new estimated signal was rebuilt though wavelet inverse transform by using the calculated wavelet coefficients.The iterative process would be continued until the terminated condition was accessed.The processing results of the ultrasonic signals with flaw from aluminum alloy forge indicates that compared to the other denoising methods based on the standard hard thresholding and soft thresholding functions,the adaptive denoising method has better denoising performance on the microstructure noise and can enhance the flaw echo signal.