Research On Home Status Recognition Based On Audio Signals

In recent years, the audio signals recognition researches has caused many scholars’ attention, especially in the speech recognition. With the development of embedded system application, the realization of speech recognition makes human being more comfortable and intelligentzed. And it has been widely used in many areas. Speech is one of the most important media for human-machine interaction, which is convenient and natural.Currently, home environmental monitoring mainly focuses on home security when no one at home. Meanwhile there are great demands on accident or anomalies alarm when old people or physical disabilities stay in alone. Video monitor system widely used in offices is normally unacceptable at home because of privacy. To solve the problem, this paper provides a real time monitoring method for home environmental anomalies based on speech identification theories. Firstly this paper analyses the basis knowledge of audio signals, the method of signal processing and introduces the history of speech recognition technology at home and abroad, its present situations and its level of researches. Based on the background of the research, this paper puts forward a research programme. The audio signals will be used as the research object in the research programme. The design of this paper is divided into two parts:the algorithm simulation and the algorithm optimization. The MATLAB7.0.4is used for developing instrument. In the part of the algorithm simulation, Hidden Markov Model algorithm and Gaussian Mixture Model algorithm are emphasized, which include:pretreatment、extracting characteristic parameters, training reference model and recognition. In the part of the algorithm optimization, the adjusted algorithm has been transplanted to the DSP development board, whose main processor is TMS320VC5502. With the help of the MATLAB Link for CCS Development Tools, the MATLAB, DSP development board and CCS software are connected to carry out simulation, debugging to complete the design. In this selection, the code is using C language to realize the function of this recognition system. Because of the CCSLink, the design can not only reduce time and cost, but also improve the reliability of the development. Basing on the above details, this paper gives a new method which has the advantages of CCS development environment and MATLAB development environment.