Abstract: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.