自平衡机器人视觉语音双重引导运动控制研究  

Research on Motion Control of Self Balancing Robot Guided by Vision and Speech

作　　者：黎小巨[1] 陈洵凛 殷素峰[1] 吴柏霖 谢小鹏 LI Xiao-ju;CHEN Xun-lin;YIN Su-feng;WU bo-lin;XIE Xiao-peng(School of Intelligent Manufacturing,City College of Dongguan University of Technology,Dongguan Guangdong 523419,China)

机构地区：[1]东莞理工学院城市学院智能制造学院,广东东莞523419

出　　处：《组合机床与自动化加工技术》2020年第10期131-135,共5页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：广东省重大科研青年创新人才项目(2017KQNCX256);东莞市社会科技发展项目(20185071511518、20185071511505);东莞理工学院城市学院重点培育项目(2018YZD001Z);东莞理工学院城市学院青年基金项目(2017QJY001Z,2017QJY004Z)。

摘　　要：自平衡机器人是一种特殊的移动机器人,为了拓展其交互能力,设计了基于视觉和语音的双重技术引导控制系统。针对目前视觉测距计算繁杂性,提出了目标定位的新方法。通过视觉测距方法,实现机器人对运动目标的动态定位;采用Camshitf目标搜索算法,在复杂环境下对目标进行连续捕获,联合双重引导控制策略,实时切换机器人的识别模式,实现机器人对指定目标的识别和跟踪,且能在预设的道路上进行自主移动。通过目标颜色、形状识别和跟踪以及视觉巡线等实验测试,结果表明该机器人交互能力强,系统反应快,视觉定位计算便捷、实时性好。Self balancing robot is a kind of special mobile robot.In order to expand its interactive ability,a dual technology guidance control system based on vision and voice is designed.In view of the complexity of current visual ranging calculation,a new method of target location is proposed.Through the method of vision ranging,the robot can locate the moving target dynamically.Camshitf target search algorithm is used to continuously capture the target in complex environment.Combined with dual guidance control strategy,the recognition mode of the robot is switched in real time to realize the recognition and tracking of the specified target,and the robot can move autonomously on the preset road.Through the experiments of target color,shape recognition and tracking,as well as visual inspection,the results show that the robot has strong interactive ability,fast system response,convenient visual positioning calculation and good real-time performance.

关 键 词：自平衡机器人 视觉定位 语音识别 目标跟踪 图像处理 运动控制 

分 类 号：TH165[机械工程—机械制造及自动化] TG659[金属学及工艺—金属切削加工及机床]