Vision-Based Legged Robot Path Planning And Data Transmission

In recent years,due to the promotion of artificial intelligence,legged robots have developed rapidly.Vision-based legged robot path planning as an important link to ensure the autonomous movement of legged robots is the focus of current research.At present,there are few mature applications of vision-based robot path planning and data transmission in the industrial field,and most of them are in the theoretical research stage of the laboratory.The thesis takes vision-based robot path planning and its data transmission system as the research object,and is committed to improving the engineering applicability of robot vision path planning.The paper builds a vision-based path planning and data transmission software and hardware platform for the robot based on the hexapod bionic robot's movement requirements,so that it can move safely and autonomously.The main research contents are: environmental data acquisition,obstacle recognition,robot path planning and data transmission.Firstly,the environment data acquisition and obstacle recognition as the front module of path planning are studied.The environment information is collected by zed depth camera,and the data is published in the robot operating system ROS as topic,and the depth image topic is subscribed in ROS for data processing.Based on Open CV,an open source computer vision library,an obstacle recognition algorithm flow is proposed,and the rationality and accuracy of the process are verified by experimental results.Secondly,based on the research of traditional path planning algorithms such as A*,Dijkstra,LPA*,etc.,the path planning of A* algorithm is implemented,and the optimization of A* algorithm has many search points and low efficiency.The fusion method of A* algorithm not only guarantees the efficiency of path planning but also improves the smoothness of the resulting path.Finally,two methods of ROS-based multi-machine communication and Wi Fi-based TCP/IP network model are used to wirelessly transmit the obstacle recognition and path planning results to the remote end for observation.Experiments show that the multi-machine communication based on ROS has a slow transmission speed,and the video is stuttered seriously;the TCP/IP network model based on Wi Fi achieves a faster data transmission speed,which can satisfy the real-time feedback of the path planning results to the remote when the robot is outdoors.Requirements.