嵌入式语音合成技术研究

发布时间：2018-02-09 02:50

本文关键词： 语音合成文语转换语音库语音索引模块语音播放模块　出处：《北方工业大学》2012年硕士论文　论文类型：学位论文

【摘要】：语音合成是将文字信息转化为标准流畅的语音朗读出来的信息处理技术。语音合成技术自提出以来已经有百年历史。经过百余年来的研究和发展,语音合成技术在合成理念、合成算法和可实现性等方面取得了巨大的进步。该技术在人机交互,文字信息处理领域有着广泛的应用。本文首先介绍了语音合成技术的发展及现状。其次对语音合成技术中的直接模拟发声法、共振峰语音合成、LPC合成、PSOLA等算法进行了系统的对比和分析研究。然后介绍了汉语语音知识和文本内容标准化的处理方法及流程。最后设计并实现了一种适于嵌入式系统上运行的文语转换系统。本文详述了该文语转换系统语音库的建立过程,包括语音单元的选择、多音字的处理、语音单元的连接、以及语音索引模块和语音播放模块的建立。本文以C++编程语言建立了语音库、语音索引动态链接库、语音播放动态链接库,并以此为基础在不调用第三方组组件的情况下,实现了一个文本语音转换应用系统,具备文本到语音转换所需的基本功能。
[Abstract]:Speech synthesis is a kind of information processing technology that converts text information into standard and fluent speech reading. Speech synthesis technology has a history of one hundred years since it was put forward. After more than 100 years of research and development, speech synthesis technology is in the concept of synthesis. Great progress has been made in composition algorithm and realizability. This technology has been widely used in the field of human-computer interaction and word information processing. This paper first introduces the development and present situation of speech synthesis technology. The resonance peak speech synthesis / LPC synthesis algorithm PSOLA is compared and analyzed systematically. Then, the processing method and flow chart of standardization of Chinese phonetic knowledge and text content are introduced. Finally, a suitable embedding method is designed and implemented. In this paper, the establishment process of the speech corpus of the speech conversion system is described in detail. It includes the choice of speech unit, the processing of multi-tone word, the connection of speech unit, and the establishment of speech index module and speech playing module. In this paper, we set up a speech base, a speech index dynamic link library and a speech playback dynamic link library based on C programming language. Based on this, a text voice conversion application system is implemented without calling the third-party group components. Basic functions required for text-to-speech conversion.
【学位授予单位】：北方工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TN912.33;TP368.1

【参考文献】