摘要
根据加性噪声的特征,应用信号复杂度特征估计信号中的噪声形态,提出一种自适应小波阈值选择方法,在小波阈值消噪中准确估计噪声阈值水平。改进了阈值消噪方法原理,使在消除信号噪声的同时保留信号中的弱特征成分。提出阈值估计的经验公式,通过理论和实践证明了这种方法能够有效地消除信号中的噪声,提取信号中的弱特征成分,并获得最大信噪比。
A new adaptive-threshold denoising method is presented. According to the method, the complexity of the signal is used to estimate the characteristic of the noise. The theory of the Donoho's threshold method is also improved. With the new method, thresholds related to the scales are estimated. Practically, based on an experiential expression, the new de-noising method can hold the features of weak components and compress the noise of the signal with optimal signal-to-noise ratio.
出处
《信号处理》
CSCD
2002年第6期509-512,共4页
Journal of Signal Processing
关键词
自适应阈值
小波
消噪方法
信号处理
Threshold Denoise Complexity Wavelet Signal processing
