基于阈值法的小波去噪算法研究

Wavelet Denoising Based on Technique of Threshold

去噪方法是信号处理中讨论得最为广泛的问题之一。人们根据噪声的统计特征和频谱分布的规律,开发了多种多样的信号去噪方法。小波变换由于具有时频局部化特点及小波基选择的灵活性,为信号的去噪提供了新的解决途径。介绍了传统的硬阈值和软阈值方法及新的小波阈值处理方法,能有效地弥补硬、软阈值方法的不足,是硬、软阈值方法很好的一个改进方案。克服了采用硬阈值法去噪效果不佳和软阈值法过度光滑使信号失真的缺点。并以实例进行比较,展现出改进方法优越性。

The method of denoising is one of the most widely discussed topics in signal processing. Based on the statistical characters of the noise and distribution rule of spectnam, various kinds of denoising methods have been developed. Wavelet transform shares some characters with time- frequency transform and it also has the flexibility of wavelet transform, so provides a new way to solve the noise problem in signal processing. The article introduces traditional threshold treatment and compares it with improved one in order to show the superiority of the advanced.