Multi-Sensor Fusion System Based on Factor Graph Optimization
- Resource Type
- Conference
- Authors
- Cheng, Xiang; Peng, Xiafu; Wu, Dongjie; Zhang, Xiaoli
- Source
- 2023 42nd Chinese Control Conference (CCC) Chinese Control Conference (CCC), 2023 42nd. :3798-3803 Jul, 2023
- Subject
- Computing and Processing
Robotics and Control Systems
Signal Processing and Analysis
Transportation
Measurement units
Inertial navigation
Cameras
Data mining
Optimization
Global Positioning System
Graph-based optimization
SLAM
Direct method
Multi-sensor fusion
Initialization
- Language
- ISSN
- 1934-1768
This paper shows a weak tightly-coupled multi-sensor fusion system, which can estimate the position and attitude of the system by combining the pixel information extracted by direct method of camera, the data of Inertial Measurement Unit (IMU) and Global Positioning System (GPS) position information. The full system constructs each module with the method of nonlinear optimization. The system determines whether the fusion is successful or not by calculating the error of each fusion module. Moreover, an incremental constraint in factor graph has been used to initialize GPS fusion effectively. The result of our experiment in different outdoor environments demonstrated the performance of our system in the accuracy.