Threshold limit values

Threshold de-noising method based on wavelet transform is an efficient method to reduce the white noise in the digital signal.

该文方法明确提出了基于小波变换的门限去噪算法中分解层数和门限阈值2个重要参数的确定方法,增强了这种去噪算法在工程应用中的实用性。

In this method,the determination of the threshold,the decomposition order and the threshold estimation model are three key problems that need to be solved.

其中,分解层数、门限阈值以及阈值函数的选取是关键。

According to the characteristics of random noise wavelet transform on the different scale and the relationship between the noise Lipschitz and its wavelet transform modulus maximal,an algorithm was present that wavelet transform coefficients of noise speech signal are filtered by changing threshold on the different scale in order to reduce noise in reconstruct speech signal.

在分析了随机噪声的子波变换系数在不同尺度上的传递特性和噪声信号奇异性与子波模极大值的关系后 ,提出了用一个尺度间变化的门限阈值来抑制带噪语音信号在不同尺度上噪声子波系数 ,从而实现了在重构信号中消除噪声的目的 。