设计了中型组足球机器人的比赛中应用的人机语音交互系统。首先简要介绍了相关软件,对语音识别原理进行了解析;然后分析了语音合成技术及其实现步骤,并根据足球机器人在比赛中的实际需要,设计了一套语音指令;利用Kinect软件进行了实验研究;最终针对不同的发声对象测试了多组数据,实验结果表明所设计的语音交互系统对语音指令的识别行之有效,识别率较高。无论是裁判机还是队员机,都能快速准确地识别队员的语音指令并进行播报,完全满足人机对抗赛对人机语音交互的需求。 The human-computer interaction system designed for the mid-sized soccer robot competition is designed. Firstly, the related software is introduced briefly, and the principle of speech recognition is analyzed. Then, the speech synthesis technology and its implementation steps are analyzed. A set of speech instructions is designed according to the actual needs of the soccer robot in the game. Experiments with Kinect software Finally, the multi-sets of data are tested for different objects. The experimental results show that the designed voice interaction system is effective in recognizing voice commands and has high recognition rate. Whether it is a referee machine or a player, can quickly and accurately identify the player’s voice command and broadcast, fully meet the man-machine competition on human-machine voice interaction needs.