Design Of Intelligent Mobile System Based On Machine Vision And The Research Of Its Obstacle Avoidance Navigation Algorithm

With the rapid development of modern science and technology,the intelligent mobile system is becoming increasingly widely used.The intelligent mobile system refers to the system which,in the absence of man-made control,can perceive the current environment through their own sensors,avoid obstacles,and complete the autonomous navigation.It can effectively complete high-risk and high-intensity operational tasks,save industrial production costs,and improve labor productivity.As the research hotspots,obstacle avoidance and navigation are the bases of stability and security of the system.Thus,the study of obstacle avoidance and navigation has essential practical value.Based on the disadvantages that the AGV(Automated Guided Vehicle,AGV)intelligent mobile system are expensive and that these devices have poor working flexibility,this paper designs an intelligent mobile system based on machine vision,focuses on researching the keys of visual obstacle avoidance navigation road image processing algorithm and obstacle avoidance navigation control algorithm,and improves the existing problems of the algorithms.The designed intelligent mobile system is operated in the simulation factory workshop environment in the laboratory,to verify that the improved algorithms can realize fast and accurate barrier avoidance navigation.The main contents of this paper include:(1)In view of the disadvantages that the electromagnetic track,laser,radar,and other equipment used by the obstacle avoidance navigation of the AGV intelligent mobile system are expensive and that these devices have poor working flexibility,this paper designs the intelligent mobile system based on machine vision,studies and improves the visual barrier navigation technology,to realize higher efficiency of the system in the workshop avoidance navigation.The overall design scheme of the intelligent mobile system is put forward,and the preprocessing technology of road image,navigation and obstacle avoidance technology are studied deeply.(2)This paper conducts detailed design of hardware circuit of the system.Selection analysis of various functional modules is made against the requirements of hardware circuit and the specific functions of each hardware module.The designed system uses STM32 as the core controller chip and uses the NVIDIA Jetson TK1 development board for processing road images.It is equipped with peripheral hardware circuits,which include one power supply module,one image acquisition module,one motor drive module,and one servo drive module.This paper also designs specific circuit diagrams.(3)This paper conducts thorough analysis and deep research on the problems of intelligent mobile system trajectory,obstacle edge detection technology,obstacle detection technology,and obstacle avoidance navigation technology,and proposes an improved scheme for the existing technology.1.The preprocessing of the road image is carried out.Then,the improved ant colony algorithm is introduced in view of the shortcomings of the traditional ant colony algorithm in the trajectory and obstacle edge detection stage.2.In the obstacle detection stage,the integrated algorithm of integral projection and region growth is proposed against the disadvantage of the large amount of calculation of the existing binocular vision algorithm.3.The integral projection algorithm is used to generally locate the obstacles,and the region growth algorithm to precisely locate the obstacles,so as to improve the efficiency and accuracy of detection.4.Against the shortcoming of the high complexity of the existing obstacle avoidance navigation strategy,the control algorithm based on data fusion are proposed to reduce the offset angle input error by complementary filtering data fusion.the avoidance PID are used in parallel to realize fast and accurate avoidance navigation of the system.(4)According to the hardware and software design,this paper builds an intelligent mobile system,and tests and analyzes the functional modules and improved algorithms.First,the modules of the system hardware are tested,and then the simulations and tests are conducted for the improved trajectory and obstacle edge detection algorithm,trajectory extraction algorithm,obstacle detection algorithm,complementary filtering algorithm,and obstacle avoidance navigation control algorithm,to verify the correct design of hardware and software and their reliability.Finally,the system is operated in the simulation factory workshop environment in the laboratory,to verify the effect of improved algorithm.