基于独立成分分析的强背景噪声去噪方法

由于许多传统的去噪方法在强背景噪声情况下提取声音信号的能力变弱甚至失效，提出应用独立成分分析(ICA)方法对声音信号进行特征提取，并证明了这种ICA变换能增强语音和音乐信号的超高斯性．在此基础上，应用ICA基函数作为滤波器，通过阈值化的去噪方法对含有强高斯背景噪声的声音信号进行去噪仿真实验．结果表明，本方法明显优于传统的均值滤波和小波去噪方法，为强背景噪声下弱信号的检测提供了新的途径．

De-noising in Intensive Noises Based on Independent Component Analysis(ICA)

As many traditional de-noising methods fail in the intensive noises environment, a method based on independent component analysis(ICA) feature extraction was applied for acoustic signals. It is proved that the ICA transform can improve the sup-Gaussian property of speech and music signals.Using the ICA basis functions as filters, a de-noising technique of threshold from intensive noises was proposed. The de-nosing experiments of acoustic signals with intensive Gaussian noise were compared with the mean and wavelet filters, the results show that the proposed method is more efficient and provides a new approach for the detecting of weak signals from the intensive noises environment.