A Vision/Map Attitude Matrix Aided IMU/Odometer Integrated Navigation Method

In the environment of complex urban roads, problems such as GNSS signal occlusion and multi-path effects caused by tall buildings, overpasses, tunnels, and dense tree canopies have brought great technical challenges to the navigation and positioning system of autonomous vehicles. In the GNSS-denied environment (such as signal occlusion and interference), the accuracy of the integrated navigation system will drop rapidly due to incomplete odometer measurement information. To solve this problem, we proposes a vision/map attitude matrix aided IMU/odometer integrated navigation system scheme. The scheme first obtains the direction, slope, and cross slope angle information of the road through positioning and high-precision maps, and constructs a road attitude matrix. Then, the scheme uses computer vision technology to calculate the angle between the vehicle and the lane line, and then obtains the vehicle attitude matrix based on vision/map information, realizing the decoupling of the odometer measurement information and the IMU attitude matrix. Based on the factor graph model, we implemented the above-mentioned multi-sensor fusion positioning algorithm and used the Matlab simulation platform to verify the performance of the algorithm. The experimental results show that our algorithm achieves a significant improvement in the accuracy of the IMU/odometer integrated navigation system in a GNSS-denied environment, which will help to promote the engineering and application of autonomous driving technology in complex environments.