Embedded Speech Recognition System Design and Optimization

Abstract

The speech recognition implementation in embedded systems is influenced by many factors. In embedded systems environment, computing rate is relatively lower, storage space is limited. Although the speech recognition technology has won some achievement in high performance platform, it's necessary to do further analysis and research on speech recognition in embedded platform. Speech recognition hardware system is constructed and high performance processor S3C2440 is taken as the key processing unit. Peripheral devices are also described in detail. In order to realize speech recognition algorithm, hidden Markov algorithm is investigated. Then embedded speech recognition software system is constructed based on Linux. After testing validation, this system has high recognition speed and accurate recognition rate, which realizes the speech recognition function. The system has good feasibility and practicability. According to the specific requirements of practical application, it can be used in vehicle terminal system, smart home system and entrance guard system, etc.