Study On Indoor AGV Autonomous Tracking Based On Vision

With the improvement of automation level,Automated Guided Vehicles(AGV)are more and more applied to production and life,and play an increasingly important role.The application of AGV can not only reduce labor costs,but also improve production efficiency.The visually guided AGV has the advantages of being free from the interference of electric and magnetic signals and easy to modify the route.In recent years,with the rapid development of computer technology and the improvement of sensor hardware capability,the visual-guided AGV has gradually become a research hotspot.In this paper,an automobile parts production workshop as an application scene,aiming at the problems existing in the actual work process,such as uneven illumination,guide ribbon damage,obstacles blocking and so on,the autonomous tracking of visual AGV was studied.The main research contents and results are as follows:1.Determine the mechanical structure of AGV and calibrate and register the camera.Based on the analysis of the advantages and disadvantages of the existing AGV drive form and gear train layout,the final AGV drive form and gear train layout were determined based on the actual working scene of the AGV in this topic,and a three-dimensional model was established Kinect camera was selected to meet the requirements,and the RGB camera and infrared camera of the camera were calibrated and registered.2.Generate the guidance trajectory with the guidance image.Aiming at the problems of uneven illumination and damaged guide ribbon in the process of work,this paper proposes a method of processing guide image,and successfully extracts the guide track from the guide image.The experiment proves that this method can effectively solve the above problems and extract the guide track effectively.3.Research on tracking control and obstacle detection.Through comparative analysis and combined with this topic design goal,choose suitable tracking control method,on the basis of kinematics analysis was carried out on the AGV,the existing pure tracking algorithm based on the improvement of AGV motion model,and use the simulation experiment of initial deviation respectively 5 cm,to 10 cm and 15 cm tracking effects of three kinds of working conditions of the simulation analysis,The effectiveness of the improved algorithm is verified.Based on the depth data obtained by the depth camera,an obstacle detection rule is designed,and an obstacle detection method is designed.The experiment shows that the method can identify the obstacle effectively.4.Experimental platform construction and algorithm verification.The AGV vehicle test platform was built,and four kinds of guided trajectory extraction experiments were carried out in the actual work process under two extreme environments,good illumination at noon and bad illumination at night,to verify the stability of the trajectory extraction scheme.The tracking capability under three working conditions with initial deviation of5 cm,10cm and 15 cm was verified by real vehicle,and the stability and effectiveness of the tracking control method were verified.Six kinds of obstacle detection experiments are carried out for large and small obstacles respectively,and the reliability of obstacle detection method is verified.