In order to solve the visual SLAM technology when applied to autonomous driving position facing the accumulation of drift problem such as larger and tracking lost, based on the closed zone autopilot visual inertial positioning system as the research object, based on the characteristics of optical flow tracking and adaptive depth estimation of visual front-end, accord with the actual motion constraints of the vehicle kinematic model was constructed. Based on the tight coupling optimization method of sliding Windows, visual inertial joint pose optimization under vehicle model constraints was realized, and a loopback detection algorithm was designed considering geometric constraints of image feature space. Finally, KITTI data set and real vehicle verification are carried out, and the results show that the proposed method can effectively improve the accuracy and robustness of the visual inertial positioning system for autonomous driving in the enclosed park, and has significant engineering application value.
Vehicle Model Constraint Based Visual-Inertial Localization Algorithm for Autonomous Driving in Closed Park
Lect. Notes Electrical Eng.
Society of Automotive Engineers (SAE)-China Congress ; 2022 ; Shanghai, China November 22, 2022 - November 24, 2022
2023-04-29
15 pages
Article/Chapter (Book)
Electronic Resource
English
Stereo Visual Inertial Odometry for Unmanned Aerial Vehicle Autonomous Flight
Springer Verlag | 2019
|Stereo Visual Inertial Odometry for Unmanned Aerial Vehicle Autonomous Flight
British Library Conference Proceedings | 2020
|