论文部分内容阅读
目的研究有限长序列的小波包处理和噪声抑制的方法．方法用正交共轭滤波器构造有限维内积空间的正交直和分解及有限长离散小波包，分析了噪声的小波包分解系数的特性，并提出小波包变换域降噪的方法．结果和结论用正弦信号和阻尼指数信号所做的降噪实验结果表明，基于最佳小波包基分解系数的硬门限处理方法能有效地抑制噪声．
Aim To study the finite length sequence of wavelet packet processing and noise suppression methods. Methods Orthogonal conjugate filters are used to construct the orthogonal decomposition and finite length discrete wavelet packet of finite dimensional inner product space. The characteristics of wavelet packet decomposition coefficients of noise are analyzed. The method of wavelet packet transform domain noise reduction is proposed. RESULTS AND CONCLUSION Noise reduction experiments using sinusoidal and damping index signals show that hard threshold processing based on the best wavelet packet decomposition coefficients can effectively suppress noise.