Research Of Embedded Speech Recognition System Based On Smart Home Control

Language is the most important tool for human communication, which is the most natural way of interaction. As a way of human-computer interaction, speech recognition is to allow the machine to be able to "understand" human language. After decades of research, speech recognition technology has been applied to the lives of ordinary people. With the improvement of living level constantly, people concept of family housing demand is changing. Smart home, the use of advanced computer, embedded systems, and network communication technology, provides safe, comfortable and pleasant high-grade family life. Speech recognition technology into smart home control via voice commands can also be for home appliances control, instead of the manual and remote control.Therefore, to enhance the research of smart home and speech recognition, and constantly improve people’s lives, has the positive significance. In recent years, with the rapid development of embedded system, embedded technology also provides solutions for the research of speech recognition and smart home. The subject of the research is based on this background, to carry out the application research of the embedded speech recognition. In this thesis, smart home control voice command, selects the method combined with the Discrete Hidden Markov model algorithm and vector quantization. By constructing an embedded hardware and software platform, the system designed speech recognition applications. The main contents of this thesis are as follows:First of all, this thesis briefly introduced the research background, outlined the development of speech recognition and smart home. For speech recognition and smart home applications, embedded technology provides hardware and software to achieve protection. This thesis introduced the basic principles of speech recognition and analysis of the main speech recognition algorithm, and smart home control technology.Secondly, in view of the speech recognition control of smart home, the thesis puts forward the basic control principle. Selected speech recognition algorithms are described; the thesis analyzes the choice of hardware and software platforms, as well as the overall design of the system. Chapters4and5detailed describe the composition and design of the hardware platform and software platform.Finally, using the software platform Visual2005in the PC machine, the thesis completes the design of speech recognition. By downloading to the embedded platform debugging, finally the thesis completes the realization of the embedded speech recognition. Through the relevant tests and experiments and analysis of the experimental results, the feasibility and effectiveness of the program are confirmed. Research on speech recognition in this thesis is according to the specific environment. Through the design of software and hardware, the thesis completed the development of the speech recognition application for Windows based on CE6.0system. The designed system can achieve the design requirements and objectives.