search

HKUST-Aerial-Robotics / VINS-Mono

A Robust and Versatile Monocular Visual-Inertial State Estimator
GNU General Public License v3.0
5.02k stars 2.1k forks source link

PreIntegration corrected by the change of bias in the paper????Can not understand the Jaccobian .......I think they maybe wrong. #132

Open GONGXI1994 opened 6 years ago

GONGXI1994 commented 6 years ago

2017-12-12 17 03 37

GONGXI1994 commented 6 years ago

Foster give the full derivation of the Jacobians, however your paper seems different from what he has done, and It would be very kind if you provide the reason why you can simply get the Jacobian from the F...... @shaojie @dvorak0 @WayneTimer @groundmelon @xuhao1

qintonguav commented 6 years ago

Calculate F is tough. Jacobian is directly part of F, even in Foster's formulation.

GONGXI1994 commented 6 years ago

I konw the F is constructed by the jaccobians ,but I do not think the "error state " jaccobian is the same jaccobian that used to correct the preIntegration when the bias change...... @qintonguav

gavinfan1 commented 6 years ago

I also can not understand the Jaccobian here. Can you give more information about the jaccobian ? I can not find the Foster's formulation. Can you give me the network link? @qintonguav @GONGXI1994

GONGXI1994 commented 6 years ago

paper: On-Manifold Preintegration for Real-Time Visual-Inertial Odometry

struggleforbetter commented 5 years ago

I'm also puzzled with this problem.The jaccobian should be the measurement change with the bias change. Why is it this? Are the two Jaccobians same ?

lywcn commented 4 years ago

I am also confused here. any updates?