New publication by Pan et al. (2024)

"Tightly coupled integration of monocular visual-inertial odometry and UC-PPP based on factor graph optimization in difficult urban environments" by Pan et al. (2024)

05.02.2024

Read
Number of comments

In recent years, high-precision navigation and positioning in urban environments have become a general concern. Rapid urbanization and dense construction have led to many urban canyons, tunnels, overpasses, boulevards, and other environments in highly developed cities. Global navigation satellite systems (GNSS) cannot provide users with accurate, continuous, and reliable absolute positions in such environments.

To enhance the positioning performance of multi-frequency and multi-system precise point positioning (PPP) in the difficult urban environment, Pan et al. propose a tightly coupled system of monocular visual-inertial odometry (MVIO) and uncombined PPP (UC-PPP) based on factor graph optimization. The initialization of MVIO and UC-PPP adopts a coarse-to-fine approach to correct the transformation of local and global frames online. Moreover, the sliding window and marginalization methods are adopted to retain the constraints between adjacent observations and eliminate useless observations in the window. The pedestrian and vehicle tests in urban environments verify the performance of the proposed method.

You can get more details by visiting the full article:
external page https://doi.org/10.1007/s10291-023-01586-3

paper_ChengPan — Structure of MVIO/UC-PPP tightly coupled integration based on factor graph optimization. After the system is started, it is necessary to initialize monocular vision and INS first, and then MVIO/UC-PPP. The initialization task of MVIO and PPP is completed with a coarse-to-fine method. The marginalization method is devoted to selecting and rejecting the observations for the sliding window to retain the constraints between the observations as much as possible and maintain the computation costs at a fixed level.