It seems all related to delta theta and, I guess, you replace Phi.block<3, 3>(0, 0) with the same element in [1].
This operation looks like FEJ-EKF?
The remainder elements, Phi.block<3, 3>(6, 0) and Phi.block<3, 3>(12, 0), confuse me.
Could you give more source about these codes?Is these related to observability constraint?
[1] M. Li and A. I. Mourikis, "Improving the accuracy of EKF-based visual-inertial odometry," 2012 IEEE International Conference on Robotics and Automation, 2012, pp. 828-835, doi: 10.1109/ICRA.2012.6225229.
You modify the transition matrix in
MsckfVio::processModel
1.
2.
3.
It seems all related to delta theta and, I guess, you replace
Phi.block<3, 3>(0, 0)
with the same element in [1]. This operation looks like FEJ-EKF? The remainder elements,Phi.block<3, 3>(6, 0)
andPhi.block<3, 3>(12, 0)
, confuse me.Could you give more source about these codes? Is these related to observability constraint?
[1] M. Li and A. I. Mourikis, "Improving the accuracy of EKF-based visual-inertial odometry," 2012 IEEE International Conference on Robotics and Automation, 2012, pp. 828-835, doi: 10.1109/ICRA.2012.6225229.