Research On Intelligent Robot Interactive Technology Based On Android

Nowadays, with the rapid development of electronic information technology, smart phones, as the leading intelligent terminal equipment, has gradually spread to the daily lives of people. According to information analysis, the production of Android devices has exceeded 1 billion in 2014, and it will continue to maintain double-digit growth in 2015. Clearly, Android has become the highest share of the smart phone operating system. Meanwhile, the rapid development of Android devices has led to the rapid development of closely related intelligent machine voice interaction technology, so the Android-based human-machine intelligent voice interaction technology has become an important research project of intelligent robot interaction. At present, the intelligent voice interactive technology develops quickly in our country, as well as the application of system and the integration of system are all becoming mature, for instance, the mobile IFLYTEK TTS (Text to Speech) system has been put into commercial use.Based on the Android smart voice interactive technology, this dissertation focuses on the intelligent voice recognition, supplemented with speech synthesis technology, intending to apply to the IOT (Internet of Things) under Android-based intelligent voice interactive system for research purposes. By studying specific speech recognition system and designing appropriate speech synthesis capabilities, the corresponding experimental system platform was built. After repeated testing, the good results were obtained, which meets the individual the needs of users in embedded transplantation and smart home environment.This dissertation focuses on the speech recognition algorithm based on the offline mode network, small vocabulary, isolated word recognition, as well as structures of Android-based intelligent speech recognition system. Then the characteristic features of the system architecture, which based on Android voice interactive mode in embedded environment, are analyzed. Also, this dissertation studied the specific speech recognition algorithm under real application background and analyses the complexity and the real-time of the algorithm modules. With some adaptive improvements, the voice recognition module and voice synthesis and broadcast module are designed to correspond to algorithms under the environment of IOT. The network offline speech recognition algorithm-Dynamic Time Warping (DTW) algoritlm and the Java program are used to write and modify the algorithm on PC and ARM machine debugging. By all above, the evaluation is good. In speech synthesis model, the IFLYTEK speech synthesis API is called to test and run the speech synthesis system in the ARM development board and achieve good results.The innovations of this dissertation are as follows. First, the voice recognition technology is implemented through the application of DTW algorithm, based on the network offline mode, small vocabulary, and isolated word recognition in. Second, the adaptability (complexity and accuracy) of related algorithms are analyzed and optimized in order to making it suitable for embedded applications based on Android environments. At the same time, by integrating the speech recognition and the speech synthesis technology, several modules are integrated into a complete, comprehensive and intelligent speech interaction and recognition system with the database technology and so on.