摘要
传统小波阈值去噪分为硬阈值去噪和软阈值去噪,而在去噪过程中,硬阈值函数在一些不连续点处有时会产生伪吉布斯现象;软阈值函数中估计的小波系数与信号的小波信号之间存在恒定偏差。为了去除这些现象,提出了一种新阈值函数,实验结果表明,新阈值函数消噪后的视觉特性较好,并且信噪比提高,从而说明了这种方法的有效性。
The typical method of threshold in de-noising has two kinds of ways,one is hard threshold and the other is soft. In some cases,such as on the discontinuities points,the Gibbs phenomenon will exhibit when the hard threshold function is used to remove noise of signals; and the soft threshold method also has disadvantages. In order to overcome the shortcomings,one new threshold function are presented. The results of experiment show that the vision of de-noising is better and the SNR has been increased,which indicates the methods presented in this paper are effective.
出处
《山西电子技术》
2010年第3期89-90,共2页
Shanxi Electronic Technology
关键词
超声检测
小波分析
降噪
ultrasonic testing
wavelet analysis
de-noising