基于语音识别的英语翻译器设计研究

Design of English Translator Based on Speech Recognition

为了提高英语翻译语音识别的识别精度,提出基于语音识别的英语翻译器设计方法。设计的英语翻译器分为客户端、网络传输端、流媒体服务器、语音信息处理器以及人机交互界面等。采用多阵列的声传感器实现对英语语音发声采集,构建英语语音信号模型,采用谱峰检测方法,进行英语发声语音信号滤波和特征提取,对提取的语音特征采用语义分析方法进行实时翻译,将语音信息输入到翻译器的信息处理器中。设计基于B/S架构的流媒体服务器实现英语翻译器的组网设计和知识库构造,在Hi3520D和Muduo网络库平台下实现翻译器的开发设计。实验结果表明,相较于对照语音翻译系统,所设计英语翻译器的语音识别准确率更高,其具备实际应用价值。

In order to improve the recognition accuracy of English translation speech recognition, a design method of English translator based on speech recognition is proposed.. The designed English translator is divided into client, network transmitter, streaming media server, voice information processor and human-computer interaction interface. Multi-array acoustic sensors are used to collect English pronunciation, build an English pronunciation signal model, filter and extract features of English pronunciation signals by peak detection method, translate the extracted speech features in real time by semantic analysis method, and input the speech information into the information processor of the translator. A streaming media server based on B/S architecture is designed to realize the networking design and knowledge base construction of English translators, and the development and design of translators are realized under Hi3520 D and Muduo network library platforms. The experimental results show that compared with the control speech translation system, the designed English translator has higher speech recognition accuracy, which shows that it has practical application value.