期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2016
卷号：4
期号：4
页码：6606
DOI：10.15680/IJIRCCE.2016.0404046
出版社：S&S Publications
摘要：Speech signal processing is one of important domain in digital signal processing because variety of noise signals can affect the speech signal so the end user cannot hear the original signal. Noises include quite, ISI, aliasing white noises, etc… This pape r presents removal of noise from original speech signal and analyzes the various parameters. This proposed approach is done by framing and windowing method. Initially the sound signal is preprocessed and notices the quite noise (i.e) silent noise and envel ope noise removes these noises by setting the maximum envelope level. The parameters include spectrogram, frequency spectrum, magnitude spectrum, thresholding, power spectral density for the noise removed signal. For the implementation of proposed work we use the MATLAB R2011b version software
关键词：End point detection; silence removal; Framing; Windowing
