Research On Integrated Navigation Technology Of Artificial Landmark Assisted INS

With the rapid development of technology,people's requirements for navigation and location services are also getting higher and higher.Whether in indoor or outdoor environment,a single navigation and positioning method is difficult to meet the needs of long term,high precision,real-time positioning.Combing two or more navigation technologies effectively through data fusion algorithms realizes a combined navigation and positioning technology in which a variety of navigation information is fused,which has become an inevitable trend of research and application of navigation and positioning technology in various environments.In this paper,the research object is the monocular visual artificial landmark aided inertial navigation system(INS).The basic working principle of the inertial navigation system,the visual navigation technology,the design of the artificial landmark,the calibration of the monocular visual scale factor,the Kalman filter fusion algorithm etc technologies are studied.The specific work content is as follows:Firstly,the related visual navigation and positioning technology and the basic principle of inertial navigation and positioning system are analyzed.For the visual navigation and positioning technology,based on the introduction of the pinhole model,this paper analyze the local feature localization technology,global feature localization technology in visual positioning technology and the vision simultaneous localization and mapping(VSLAM)technology as well as the landmark-based visual positioning technology.For the inertial navigation and positioning technology,the coordinate system and the motion state representation method commonly used in navigation and positioning are introduced.And how to use the output information of gyroscope and accelerometer to solve the state and position information is analyzed.Secondly,an artificial landmark is designed,and the problem of missing scale information in monocular vision is solved by calibration.The monocular visual positioning experiment based on the designed artificial landmarks is carried out.For the design part of the artificial landmark,an artificial landmark with circular peripheral contour is designed,and the round landmark in the image is detected by Hough.Shooting and calibrating the landmark with known distances effectively solves the problem of missing scale information in monocular vision.Then,using the 3D-3D inter-frame pose estimation method,based on the known landmark location information,a monocular visual artificial landmark location experiment is carried out.The experimental results show that the average positioning error is less than 5cm at a distance of 0.5m to 2m from the landmark.Finally,the combined navigation and positioning method of monocular visual artificial landmark aided INS is researched.First,we set up artificial landmarks in the known experimental scenes and build a visual landmark database based on the local navigation coordinate system.In the working process of the navigation and positioning system,the monocular visual positioning information obtained by the 3D-3D inter-frame pose estimation method is merged by the Kalman filter algorithm and the positioning data of the INS,and the pose information obtained by the monocular vision is used.Pose information obtained by monocular vision is used to correct the cumulative error of the inertial navigation.The final comparison experiment results show that the average positioning error of the combined navigation and positioning method for the monocular visual artificial landmark assisted INS is 0.154 m.Compared with the traditional dead reckoning method,the performance is improved by 78.3% and the cumulative error of the inertial navigation system is effectively reduced.