一种改进小波阈值去噪方法的研究

针对传统小波阈值去噪分析的不足，分别从两个方面进行改进：1）根据小波系数在各尺度上的相关性，提出基于Lip指数阈值寻优的新方法；2）用构造的新小波阈值函数处理小波系数，克服了软阈值的高阶不可导和硬阈值函数的不连续，且计算方便。最后使用新方法对一组离散变形数据进行去噪处理，验证了其有效性和优越性。

Improved Wavelet Threshold De-Noising Method Research

This paper improves traditional wavelet threshold de-noise analysis from two sides. First,a new method based on Lip exponent threshold optimization is proposed by the correlation of wavelet coefficients at different scales. Second, we construct a new wavelet threshold function to deal with the wavelet coefficients. This is convenient to compute, and it overcomes the discontinuity of the hard threshold function, without the high order derivative of soft thresholding function. Finally, we process a set of discrete deformation monitoring data using the improved threshold algorithm method, showing its better de-noising ability.