Development Of Embedded Speech Recognition System And Its Application Research In Smart Home

With the progress of human society and the rapid development of science, people began to pursue smart and convenient living environment, so smart home quickly rises. The significance lies in its smart home facilities can humanlike thinking or can understand the intention, voice control is the best choice for smart home control to change the traditional key-contact operation. Although speech recognition technology in high-performance platform has already got some performances, due to the embedded system environment is relatively low computing speed, its storage space is limited, and the real-time speech recognition have certain requirements, but it is necessary doing further research for speech recognition technology in embedded platforms. If speech recognition can be applied to the embedded platforms, it will broaden application prospects, and bring great economic benefits.The project take the smart home as a backdrop, carrying out the research of speech recognition system, select the method combining semi continuous hidden markov model algorithm and vector quantization for speech recognition to control household equipment, and establish a smart, convenient home intelligent control system, to make life more intelligent. The main work and results are as follows:1. The paper introduced the research background and significance of this topic, and overviews smart home, research status of speech recognition at home and abroad and embedded speech recognition technology. This paper introduces the speech characteristic parameter and the speech recognition algorithm commonly used, put forward in the speech recognition system applied to embedded platform design principles.2. The improvement of speech recognition algorithm is elaborated, the design scheme of the hardware, software is analyzed, and determines the overall implementation of the system scheme. This paper explains the system hardware circuit design and software programming in great detail.3. The compiled program is downloaded to the hardware platform debugging, finished the development of embedded speech recognition system, which is applied to smart home control system. Through the relevant testing and practical application scene, the feasibility of the system design scheme and the practical of household equipment controlling was confirmed.Embedded speech recognition system in this paper is developed for the smart home application scenarios, completed the development of a speech recognition system, and applied to the control system of a smart home, achieved the system design requirements and objectives.