Visual Inertial Navigation Fusion Location And Integrity System Design

In recent years,people's dependence on location services has become more and more high,Although the global satellite positioning systems has provided most location services,satellite signals are seriously disturbed indoors due to the blockage of building walls and glass is still a big problem.In some indoor places,such as shopping malls and underground parking garages often provide not only the correct location services,but sometimes even misleading people's location information.Both the visual positioning system and the inertial navigation system can provide location services in indoor environments.This paper uses a visual and inertial design fusion positioning system to provide indoor location services for pedestrians.Firstly,investigates the theory and development of visual odometer and visxual inertia odometer,compares the different ways of fusion of vision and inertial navigation,and then studies the mainstream visual inertia odometer project at home and abroad.Secondly,aiming at the needs of pedestrians in indoor positioning and their scene characteristics,a visual inertial fusion system is designed and the scalability and stability of the system are studied.The designed positioning system adopts modular design.According to the function s of each part of the system,it is divided into visual module,positioning subsystem module,fusion module and correction module,and different implementation schemes are designed for different principles and requirements of each module.Then,the integrity of each module and the influencing factors of the visual module are analyzed.The factors such as the light intensity,the different illumination of the left and right cameras,the synchronization of the left and right cameras,the movement process of the pedestrian,and the moving objects in the field of view are considered.Exploring the operation of the vision module under different influencing factors,and designing a corresponding state anomaly solution to ensure effective monitoring and control of the state of the vision module.Then,using the improved particle filter map matching algorithm to correct the error of the system,which improves the positioning result and stability of the system.Finally,the experiment to analyzes and tests the integrity of the system is designed in this paper,and compares the positioning of the inertial navigation data with pure visual positioning,visual fusion and error,and the positioning result of visual fusion high-precision inertial navigation data,which proves the proposed system has a better performence on the feasibility and effectiveness in indoor location service.The visual inertial navigation and positioning system designed in this paper can improve the stability and accuracy of indoor positioning,which can provide a good location service for indoor environments such as large shopping malls and underground garages.