中文摘要：

阈值和阈值函数的选取是小波去噪的关键。经小波分解变换后,低频部分包含大量的有用信号,而噪声分布于整个小波域内的高频部分,其中包含少数有用信号的细节信息。采用固定阈值和阈值函数处理时会造成高频部分的有用信号细节信息丢失。基于邻域小波变换将分解层数引入阈值和阈值函数中,使用动态阈值和阈值函数处理不同分解层数下的小波系数,来保留不同层数下的细节信息。仿真结果表明所提方法的信噪比和信噪比增益最大,且均方差最小,相关系数逼近于1,原始信号与重构信号相似度得到提高,改善了去噪效果。

英文摘要：

Appropriate selection of threshold and threshold function is the key problem of wavelet de-noising. After the wavelet transform, a large number of useful signals are contained in the low frequency part, but the noise will distribute in the high frequency part, in which the detailed information of a few useful signals is contained. Even though dealt with fixed threshold and threshold function, some detailed information in high frequency will miss. Therefore, the decomposition scale is introduced into the threshold and threshold function based on neighborhood wavelet transform, using dynamic threshold and threshold function to cope with the wavelet coefficients in different decomposition scales so as to researve the details of different scales. The simulations reveal that the proposed method can not only make the signal-to-noise ratio and the gain of signal-to-noise ratio maximum, but also obtain the minimum root mean square error. Meanwhile, the correlation coefficient approaches 1. The similarity between original signals and reconstruction signals is improved, which implies that the de-noising effect gets better.