Research On Path Identification Technology Of The Mobile Robot Based On Machine Vision

With the rising labor cost, the traditional methods of goods-taking are eliminated gradually. Most of the automatic guided vehicles goods-taking method use fixed magnets technology. For this method, the route choice is single, adaptation ability is weak and can't ensure the real-time path recognition and tracking, so the delivery time is delayed.In recent years, with the continuous development of machine vision, using machine vision to recognize and track the path has becomes a research hotspot. A path recognition system based on machine vision is studied in order to solve the problems of goods-taking.Firstly, the “QR codes and machine vision” technology was used based on the theory of robot positioning technology. The road location points of the environment coordinates in the warehouse were obtained by establishing grid modeling of the warehouse. The road location point was encoded using the binary number and the QR code was posted in the location of point. The accurate position was recognized by processing QR code using machine vision. The robot control system can use this information to positioning precisely. The method can improve the recognition ability of the robot, shorten the location time and improve the efficiency of the system.Secondly, based on the established static environment warehouse model, through the comparison and analysis of various path planning algorithms, the global path planning of the initial static warehouse was finished by a kind of genetic algorithm, the improved A*algorithm is used to optimize the local optimization. This method successfully solved the problem of blind search path, shortened the time needed for path searching and satisfied the real-time requirements of robot.Finally, the image of the road line with red mark was collected by the machine vision and processed, and the least square method was used to fit the center of the red marking line through using the machine vision technology. The path following model was established according to the relationship between the fitted line and the robot pose.The robot was controlled in real-time using fuzzy control technology to make sure it travels along the established route, and complete the tracking of the established route.In the process of experimental design and research, the accurate localization and tracking of the established route were achieved and approved by the robot field test of the system, the robot can reach the target point effectively and accurately, It is proved that this system has practical and feasible use for the identification of the path in the warehouse.