The Research And Implement Of Embedded Speech Maps Based On SPCE061A

Currently, the navigation system has gradually become a hot topic. There are various navigation products in the market, including on-board navigation systems, ocean navigation systems. However, most of the navigation system does not yet have all speech recognition functions, such as PDA,PPC, Smart Phone and other consumer electronic navigation system is the lack of human-computer interaction, the most simple, The most direct forms -- voice interactive.To address these issues so that the navigation system more efficient and effective. The goal of this project is the introduction of speech recognition, and navigation systems integration, achieve a complete set of voice recognition navigation system. Expand the use of voice recognition, rich navigation system's function, do a real man-machine communication. Research is focused on early speech recognition part of the design and implementation.This paper introduces embedded speech recognition technology status and the future development trend. Explore the problems Embedded Speech Recognition facing. Study the main flow of current speech recognition and the current typical speech recognition model matching algorithm. Detailed analysis of the speech recognition model features, and then from the hardware platform and software structures design sides, Development of a comprehensive speech recognition system. In which the area of software design, focused on the interruption of the system, serial communication, the code system conversion, and other core issues.In this paper the characteristics between the following points : (1) Using MATLAB detailed analysis of the various voice recognition principle of steps, including voice signal preprocessing, feature extraction, pattern matching, and so on. (2) Improved Dynamic Time Warping algorithms, comparative analysis of the improved algorithm with the original algorithm. (3) Analysis of the speech signal noise on the recognition rate, Wavelet signal noise reduction upgrade the HMM algorithm to identify the success rate.Summing up, we built a new embedded speech recognition navigation system; it embedded equipment, LCD display module and serial communication module organic integration. The system with low power consumption, scalability strong advantages for the navigation system of opening up new prospects.