智能玩具的语音识别系统.docVIP

PAGE 摘 要 目前，语音这种人类最简单、最方便的信息交换方式，在玩具上的应用还较少。显然，在玩具系统上增加语音接口，用语音代替键盘输入，并进行人机对话，让玩具能听会说，不仅十分诱人，而且具有广泛的应用前景。而语音识别技术是近年来十分活跃的研究领域，语音识别系统的实用化研究是语音识别研究的一个主要方向。 用单片机控制语音芯片，再把单片机和语音芯片嵌入到玩具中就可以做成语音播放的玩具了。本语音识别系统以LD3320语音识别芯片为核心部件，MCU选STC10L08XE。通过麦克风将声音信息输入到LD3320进行频谱分析，将提取到的语音特征和关键词语进行对比匹配，找出相似度最高的关键词语作为识别结果输出给MCU。MCU通过读取SPI Flash里的语音数据播放相应的语音作为回应。 本系统由语音识别模块，电源模块，程序下载模块，MCU模块，语音数据储存模块组成。论文概述了各个模块的工作原理，并对电路的可靠性、程序的可行性进行了论证。 系统给人以比较高的用户体验，能够稳定的工作，该方案结构简单，具有较高的应用价值。 关键词: 语音识别； LD3320； STC10L08XE单片机 Abstract At present, the human voice, the simplest and the most convenient information exchange ways, in the application of the toys is less. Obviously, if we add voice interface in the toy system, instead of the keyboard, let toys can listen and said, this is not only very appealing, but also has a broad prospect of application. The speech recognition technology is a very active field of research in recent years. The application of the speech recognition system is the major research direction. With single-chip microcomputer controlling voice chip, and then put the single-chip microcomputer and voice chip embedded in toys we can make speech broadcast toys. This speech recognition system make LD3320 speech recognition chips as the core components, and we choose STC10L08XE as the MCU. Through the microphones, sound input to LD3320 for spectrum analysis. Phonetic features will be extracted and LD3320 trys to match the key words, finding out the similarity of the highest key words as recognition results and transmit it to MCU. The MCU reads the speech data in the SPI Flash to play as a response to the speech. The system is composed of speech recognition module, power supply module, the program download module, MCU module, and voice data storage module. This paper summarizes the principle of work of the modules, and the reliability of the circuit, and has proved the feasibility of the program. That the system gives a person with higher user experience and it works stable, the sch