Introduction: This project is used for an online course about visual inertial odometry on shenlanxueyuan. It is based on the VINS-Mono framework but does not rely on ROS, Ceres, G2o. This code is very basic and aims to demonstrate Eigen-based back-end LM algorithms, sliding window algorithms, robust kernel functions, etc. This code supports for Ubuntu or Mac OS.
- exercise1: Rotation update of Lie algebra and quaternion
- exercise2: IMU trajectory integration and IMU noise calibration
- exercise3: Use LM algorithm to fit a curve
- exercise4: Nullspace of Hessian Matrix in Bundle Adjustment
- exercise5: Use Schur Complement to accelerate computation of Hessian matrix
- exercise6: Triangulation of landmark in Mono-SLAM
- exercise7: Vins Mono on euroc dataset