Airborne Voice Recognition Technology And Environment

The research goal of speech recognition is to invent a kind of machine which has the function of hearing and can give correct response directly according to people's oral order. With the development of information science and technology advance, there is higher requirement for the technology of command and control. Speech recognition will be necessary for comprehensive electronics equipments, such as automobile, airplane, and ship. The research of speech recognition in airplane is significantly important for pilot to give all kinds of necessary orders when it is flying with high speed, in order to improve the comprehensive response speed in the battle system.In this background, the article mainly researches the speech recognition system based on the small isolated word on specific people and specific item, as well as gives detailed research on the algorithm of speech recognition system. It finally designs a speech recognition system based on DSP, focusing on ADSP-21161N from ADI Company , as well as give detailed introduction to the hardware design and software design.The article mainly involved several aspects as follows,1. Introduction to the research and development situation of speech recognition, and brief introduction on the basic knowledge of speech signal processing. Based on it, there is description on the basic theory and main method of speech recognition, which provides great foundation to the research of speech recognition。2.Detailed analysis of the signal processing procedure and relevant algorithm of speech recognition system。Detailed analysis and comparison to current method of speech endpoint detection and researching the algorithm of DTW and the model of HMM, and therefore test all kinds of these algorithms through Matlab.3. Peculiarity and architecture of the ADSP-21161N was introduced, and the speech recognition system was raised. How to design the circuit based on the ADSP-21161N was researched, and the consist of the system and how the process of the system works was discussed. Analog and digital convert is performed by TLV320AIC23, training and recognizing of speech signal are performed by DSP, all the logic control is performed by the FPGA, the result of the recognition will be output finally.4. According to the characteristics of isolated word speech recognition system focusing on specific person and specific content, it designs software through choosing the DTW algorithm, and describes and discusses the procedure of the software design, which can basically complete speech recognition algorithm.5. Based on the whole article, there is a analysis on the current problems of the researching subject, which giving a direction for further research.