• 首页
  • 期刊简介
  • 编委会
  • 投稿指南
  • 收录情况
  • 杂志订阅
  • 联系我们
引用本文:白燕燕,胡晓霞.基于MATLAB语谱图的声乐研究[J].软件工程,2019,22(9):1-4.【点击复制】
【打印本页】   【下载PDF全文】   【查看/发表评论】  【下载PDF阅读器】  
←前一篇|后一篇→ 过刊浏览
分享到: 微信 更多
基于MATLAB语谱图的声乐研究
白燕燕,胡晓霞
(西安工业大学北方信息工程学院,陕西 西安 710200)
摘 要: 语谱图是一种在语音分析以及语音合成中具有重要实用价值的时频图,能反映出语音信号动态频谱特征, 被认定是语音信号的可视语言。横坐标代表时间大小,纵坐标代表频率大小,时间和频率所对应的像素点的值表示能量值 的大小。本文对一小段语音信号和音乐信号进行处理,最终以语谱图的形式展现出来,并通过对语谱图的分析来判断语 音信号和音乐信号,检测语音是否开始和结束,效果良好。这种方法容易区分语音信号中的有用声段和噪声声段。实验 证明,纯语音信号频域能量和过零率变化要大于音乐信号。音乐信号较语音信号能量谱稳定。
关键词: 语谱图;语音分析;语音合成;短时能量;过零率
中图分类号: TP391    文献标识码: A
基金项目: 本文系西安工业大学北方信息工程学院2019年度院长科研基金资助项目(项目编号19YZ01).
Research on Vocal Music Based on MATLAB Spectrogram
BAI Yanyan,HU Xiaoxia
( Xi'an University of Technological Information, Xi'an 710200, China)
Abstract: Spectrogram is a kind of time-frequency graph with important practical value in speech analysis and speech synthesis.It can reflect the dynamic spectrum characteristics of speech signals and is regarded as the visual language of speech signals.The horizontal coordinate represents time,the vertical coordinate represents frequency,and the value of pixels corresponding to time and frequency represents the magnitude of energy value.This paper deals with a small segment of speech signal and music signal,which are finally presented in the form of spectrogram.By analyzing the spectrogram,we can judge the speech signal and music signal,and detect whether the speech starts or ends.This method is easy to distinguish the useful sound segments from the noise segments in speech signals.It is proved that the change of frequency domain energy and zero crossing rate of pure speech signal are greater than that of music signal.The energy spectrum of music signal is more stable than that of speech signal.
Keywords: spectrogram;speech analysis;speech synthesis;short-term energy;zero-crossing rate


版权所有:软件工程杂志社
地址:辽宁省沈阳市浑南区新秀街2号 邮政编码:110179
电话:0411-84767887 传真:0411-84835089 Email:semagazine@neusoft.edu.cn
备案号:辽ICP备17007376号-1
技术支持:北京勤云科技发展有限公司

用微信扫一扫

用微信扫一扫