The traditional visual navigation method for quadruped mobile robots mainly focuses on navigation monitoring,and the real-time map accuracy of visual positioning is low,which affects the visual navigation effect.To this end,a machine learning based visual navigation method for quadruped mobile robots is designed.This method is based on machine learning to extract visual features of quadruped mobile robots,extract navigation information from real-time visual images obtained by the robot,and provide three-dimensional structural information of the visual environment through machine learning analysis.The experimental results show that this method can quickly find the optimal path,complete visual navigation tasks,reduce robot energy consumption,and effectively improve overall efficiency.
关键词
机器学习/四足移动机器人/视觉导航方法
Key words
machine learning/four-legged mobile robot/visual navigation method
维普
万方数据