语音信号特征参数提取方法设计与实现毕业设计.docVIP

LANZHOU UNIVERSITY OF TECHNOLOGY 毕业设计 题 目 语音信号特征参数提取方法设计与实现 学生姓名 ** 学 号 ****** 专业班级 通信工程*班 指导教师 *** 学 院 计算机与通信学院 答辩日期 ***************** 语音信号特征参数提取方法设计与实现 Design and Implementation of Speech Signal Feature Parameter Extraction Method ** ** *** 摘 要 随着计算机技术的发展，语音交互已经成为了人机交互的必要手段。语音特征参数的精确度直接影响着语音合成的音质和语音识别的准确率。因此语音信号参数提取是语音信号处理研究中一项非常有意义的工作。 本文采用Matlab软件提取语音信号特征参数，首先对语音信号进行数字化处理，其次，进行预处理，包括预加重、加窗和分帧，本文讨论了预处理中各种参数的选择，以使信号特征提取更加准确。再次，讨论了各种时域特征参数的算法，包括短时能量、短时过零率等。 实验结果表明，通过对语音信号进行短时能量及短时过零率的分析与提取，可以初步从一段语音信号中分辨出有音还是无音，在有音状态下还可以分辨出轻音与浊音。 关键词：语音信号；特征参数的提取；短时能量；短时过零率 Abstract With the development of computer technology, voice interaction has become a necessary means of human-computer interaction, the accuracy of the speech feature parameters directly affects the quality of speech synthesis and speech recognition accuracy. So the speech signal parameters extraction of speech signal processing research is a very meaningful work. First，This paper extracts speech signal characteristic parameters ,and digital processing the speech signal by the Matlab software .second, preprocesses the signal, including preemphasis, add window and framing, this paper discussed the selection of various parameters in pretreatment, in order to make a more accurate signal feature extraction. Third, discusses the various characteristic parameters of time domain algorithms, including short-time energy and short-time zero crossing ratio, etc. Experimental results show that by short-time energy and short-time average power of speech signal, and short-time zero crossing ratio analysis and extraction from the preliminary period of speech signal distinguish with or without, and in sound condition can discern the light tone and dullness. Key words: speech signal; The extraction of feature parameters. Short-time energy; Short-time zero crossing rate目录 第一章 绪论 1 1.1 语音信号特征提