motokimura/kalman_filter_with_kitti

Applying extended Kalman filter to KITTI GPS/IMU data for vehicle localization

kalman_filter_with_kitti's Issues

R_ and self.P

Hello and congrats for your job. I want to ask two questions.

1. In the Kalman filter main routine, you do this: R_ = R * (dt ** 2.) but you dont use the R_
2. Based on the book Probabilistic Robotics, the resulting covariance in the update step is given by: Σt = (I − Kt Ht) Σ¯t (3.62)
   However in your code, is given by: self.P = self.P - K @ H @ self.P. Shouldn't be self.P = I - K @ H @ self.P?

Thank you in advance.

when i run demo.ipnb,there is a alamr: no module named kalmanfilter? what can i do? pip install pyfilter?

Hi, I would like to ask about the model for state increment matrix calculation

Hi, thanks for your work,, I would like to ask about the physical model for the state increment matrix calculation(How the increments of x,y are computed？).🙂

Theta slow convergence

Hi, thanks for the good EKF implementation. However, I noticed that the theta estimation shows a significant convergence delay. How this can be improved?

GPS Update

Hello, i would like to ask how you update the theta angle in the update step? The GPS provides you only with position measurements.
Thank you in advance for your time.