| With the advancement of technology and the gradual improvement of robot technology,intelligent robots can provide more intelligent services in place of humans.For example,the inspections of patrol robots at substations,the detection of abnormalities at specified locations,and the capture of images to local servers,or the transportation of the images back to the backend server;the museum guide robot can reach the designated cultural relics site to explain the history and culture of the cultural relics to tourists.So,how to reach the designated location of the robot to achieve video surveillance and voice services is a major problem to be solved.Robots can utilize the navigation technologies to reach a specified target point.Current mainstream robot navigation technologies include magnetic navigation,laser navigation,GPS navigation,landmark navigation,and visual navigation.In view of the characteristics of visionbased navigation,what are good flexibility,low cost and easy implementation,this article uses vision-based navigation based on the laying of navigation lines on the ground to realize the robot's autonomous navigation,self-planning path,self-positioning to achieve the target point,and in this position to perform video surveillance and voice services.The main contents of the study are briefly described as follows:(1)Autonomous navigation.First of all,the guide-line is laid for the robot,and the robot obtains guide-line map through vision to calculate the offset angle and offset distance between the guide-line and its corresponding virtual criterion-line in the map.Then,the PD(proportional-differential)fuzzy controlling rule is used to adjust the robot's straight travel speed and rotation angle,so that the robot can autonomously correct and go straight along the guide-line.(2)Autonomous positioning.It is purpose that designing a nine-square grid coding pattern at the crossroads of the guide-line to achieve the robot's node positioning.Based on the recognition principle of the WeChat two-dimensional code,three locations in the four top corner of the coding area on the nine-square grid are used to design a positioning grid,which solve the problem that identifying the path node from different directions at the crossroads keeps the result consistent.(3)Path planning.Taking advantage of the global electronic map information provided in advance,through the Dijkstra path algorithm,the robot automatically plans the optimal path from the initial point to the target point and adjust the robot by oneself to realize along the path to reach the destination.(4)Video surveillance.The robot can realize to monitor remotely the information of the running path on real-time by the network transmission technology of the computer.It can also reach the designated target point to record and capture the surrounding information,which can be saved to the local server or transmitted back to the background control terminal to save through the network.(5)Voice service.Pre-storing different textual information for different path nodes,when the robot reaches the designated node,it can broadcast the corresponding text information by voice to make a related introduction.At the same time,the robot can achieve straight,backward,turning or stopping by recognizing simple voice commands.This paper solves the problem that the robot makes use of vision-based navigation to autonomously adjust to reach the designated node,and realizes the functions such as video surveillance,capture the surrounding environment at designated locations and voice broadcast.The innovation of this paper is that the design of nine-square grid template based on the recognition principle of WeChat two-dimensional Code solves the problem that the same path node keeps consistent results from different directions. |
