Long-term navigation for aircraft in global position system (GPS) denied environment is a very challenging task. In this thesis, a novel aircraft autonomous positioning method based on fusion multi-modal image matching and inertial navigation system (INS) is proposed. This method consists of three steps. First, the coordinates of ground control points were obtained by matching the real-time image with the reference image. Then, PnP algorithm is employed for calculating the position of the aircraft. Finally, aircraft gets the optimal position by using BP neural network to combine the position estimated by inertial navigation and the results of multi-modal image matching. Simulation result displays that our method can effectively improve the positioning accuracy of aircraft. The accuracy of this method proposed in this thesis is comparable with that of GPS. Therefore, the study of this data fusion method has certain effects, the positioning accuracy can be improved of the aircraft. It is a great significance for the aircraft to achieve autonomous navigation.
Vision Inertial Fusion Based on BP Neural Network for Aircraft Autonomous Positioning
Lect. Notes Electrical Eng.
2021-10-30
11 pages
Article/Chapter (Book)
Electronic Resource
English
Vision Inertial Fusion Based on BP Neural Network for Aircraft Autonomous Positioning
British Library Conference Proceedings | 2022
|Robust Autonomous Navigation Method for High-Precision UAV Based on Inertial/Machine Vision Fusion
Springer Verlag | 2024
|Pedestrian Autonomous Positioning System Based on Inertial Navigation
British Library Conference Proceedings | 2020
|