融合尺度加权的改进小波阈值函数的超声去噪

Ultrasonic denoising based on a improved wavelet thresholding function merged with scale weighting

为了提高超声在线检测的准确度,有效去除噪声,引入了算术平均滤波方法对信号进行预处理。针对小波阈值去噪方法中传统阈值函数在阈值点处不连续、存在恒定偏差的缺点,在研究现有改进阈值函数及尺度加权方法的基础上,提出一种融合尺度加权的改进小波阈值函数。仿真结果表明:改进后的小波阈值函数可显著提高信噪比,并在超声薄板厚度在线检测的实际应用中,具有很好的去噪效果。

In order to improve the accuracy of ultrasonic online testing and to denoise efficiently, the arithmetic mean filter is used to pretreat the signal. In view of the shortcomings of traditional threshold functions in the wavelet thresholding denoising method, on the basis of studying current threshold functions and scale weighting method, an improved wavelet thresholding function merged with scale weighting coefficient is proposed. Experimental results show that the new denoising method has obvious advantages in the Signal Noise Ratio （SNR） aspect ,it also has good denoising effect in ultrasonic online testing of sheet metal in practical ap- plications.