【摘要】 为有效剔除信号噪声,得到没有污染的信号,便于进行信号分析,介绍了小波包变换的改进方法,给出了信号的小波包分解系数排序方法,将信号的相关性与小波包的多尺度时频分析及重构结合在一起,并用于信号的去噪研究中。通过仿真研究和与目前小波去噪方法进行对比,证明了该方法去噪的有效性。更多还原

【Abstract】 In order to eliminate signal noise, an improved wavelet packets transform is introduced with wavelet packets decomposing coefficient array given, and relativity of signal, multi-scale time-frequency analysis and reconstructing are put together for de-noise study. This method is proved to be effective by a simulation example and comparison with recent de-noise using wavelet theory.更多还原