语音信号特征提取算法研究

Algorithm for feature extraction of speech signal

语音作为一种搭载着特定的信息模拟信号,已成为人们社会生活中获取信息和传播信息的重要的手段。语音信号处理的目的就是在复杂的语音环境中提取有效的语音信息。环境干扰在语音传播过程中对信号的影响不容小觑,因此语音信号处理的抗噪声能力已经成为一个重要的研究方向。Matlab的应用有着广泛的领域,在信息处理领域其强大的数据处理能力可以将非平稳时变的语音数据转换为离散的数据,然后可对离散数据进行分析或者做进一步运算处理。它的信号处理工具箱可以迅速、有效地实现语音信号的处理和分析,Matlab是适用于信号处理领域的强大的处理工具。在此运用Matlab对一段包含有环境噪声的语音进行傅里叶变换、时域和频域分析、提取部分语音信号及分析信号的处理。

Speech signal processing is to extract the effective speech information from the complex speech environment. The influence of environmental interference in speech communication process on the speech signal should not be underestimated,so the anti-noise ability of speech signal processing has become an important research direction. Matlab has a powerful data processing capacity in the information processing field and can convert the nonstationary time-varying audio data into discrete data,and then performs discrete data analysis or further processing. Its signal processing toolbox can quickly and effectively achieve the processing and analysis of speech signal. It is a powerful processing tool in the signal processing field. For a piece of speech containing environment noise,Matlab was used to do Fourier transform,time domain and frequency domain analysis,extraction of partial speech signal and analysis of signal characteristic.