Research On Visual Inertial Navigation Algorithm Using Filtering Method In Lie Group Space

In recent years,mobile robot technology has developed rapidly and has been widely used in industrial,agricultural,military and other fields.Precise pose estimation of robots is the basis for autonomous navigation and intelligent control,and it is also one of the hot topics in this field.The good information complementarity between visual sensors and inertial navigation makes the pose estimation based on fusion of visual inertial navigation become the mainstream of current robot navigation.How to effectively integrate visual and inertial navigation information to obtain higher positioning accuracy is one of the main challenges in this field,and it has important research value and significance.In order to effectively convey the uncertainty of the estimation system and reduce the non-linearity error of the system,this paper studies the vision-inertial navigation fusion positioning algorithm in Lie group space,and describes the uncertainty of the system by constructing a high-dimensional Lie group matrix In this paper,the tightly coupled filtering method is used to realize the fusion of visual and inertial navigation information,and finally the precise positioning of visual inertial navigation fusion is completed.This article mainly researched from three aspects:Firstly,it introduces several attitude solution methods and their conversion relations,and outlines the composition and related principles of inertial navigation system.In the part of the inertial navigation system,the IMU model and the IMU pre-integration algorithm are analyzed and explained,which lays the foundation for subsequent research.Secondly,this paper analyzes the conversion relationship between several coordinate systems of the system,and completes the calibration of the camera and inertial guidance.In this article,Zhang Zhengyou's checkerboard calibration method was used to implement camera calibration,and the camera's internal and external parameters were obtained.The quaternion method was used to implement the camera-inertial navigation joint calibration.Finally,this paper elaborates the visual inertial navigation fusion positioning algorithm in Lie group space,and proposes an I-IEKF filtering algorithm combining iterative Kalman filtering and invariant Kalman filtering.The algorithm uses high-dimensional Lie groups to construct system state variables,and iterative methods based on invariant Kalman filtering are used to further reduce the non-linear errors of the system and improve the accuracy of pose estimation.This paper describes the algorithm process in detail,and verifies the algorithm performance in the EuRoc dataset.The results prove that the proposed algorithm is effective in improving the accuracy of visual inertial navigation fusion pose estimation.