A python implemented error-state extended Kalman Filter. Suit for learning EKF and IMU integration.

Use simulated imu data (./data/imu_noise.txt) as input.

Output an trajectory estimated by esekf (./data/traj_esekf_out.txt) and a ground truth trajectory (./data/traj_gt_out.txt).

You can use evo to show both trajectories above.

evo_traj tum ./data/traj_esekf_out.txt --ref ./data/traj_gt_out.txt -p

1. Quaternion kinematics for the error-state Kalman filter
2. State Estimation for Robotics.
3. Probabilistic Robotics.