The Research And Implement Of Isolated Word Speech Recognition Algorithm Based On DTW Model

To be a comprehensive and interdiscipline subject, now, speech recognition technology has formed well-established theoretical system after the continued research for nearly about 40 years.Now the research of speech recognition technology focuses on practical applications.It provides a typical solution of human-computer interaction for today's information society which will bring about revolutionary changes in people's lives and production.Speech recognition technology can be applied to all sectors of society. Considering the rapid development of speech recognition technology and a broad market background in the recent years, this paper designed a kind of isolated word speech recognition system to meet the need.This paper introduces the complete development process of speech recognition system.It analysis and designs the principle and algorithm of front-end preprocessing, edge detection and feature extraction and Lay heavy stress on the endpoint detection algorithm with noise, MEL frequency cepstral coefficients(MFCC)features parameter extraction algorithm.It also gives the corresponding simulation results for the design algorithm.The results satisfy the design requirements of system performance. It compares the traditional aspects of the dynamic time warping (DTW) algorithm in the part of pattern matching and raised the all path constraint DTW algorithm (ADTW), simulation shows that the algorithm can effectively improve the recognition rate and has excellent robustness form using the MATLAB language to program the speech recognition algorithm.Meanwhile, the paper designs the device of portable speech recognition algorithm pulmonary function based on microprocessor chips ARM9 S3C2440 and the algorithm of speech recognition.It details the various modules of the system hardware implementation. It designs the driver of voice collecting and the realization of speech recognition algorithm based on the transplant of WinCE5.0 in the aspect of hardware design.The results show that the speech recognition algorithm has high recognition rate, can reduce or eliminate noise caused by the training model and the mismatch between the speech test. The device of pulmonary function based on this algorithm fully meet the design requirements.It will have a very good market prospects.