There are many resources that relate to EKFSLAM based on the presence of only a LiDAR unit, which localize both the vehicle and the detected landmarks.

I see that they often use state vectors that get augmented as new unique landmarks are detected, in the form of:

\begin{pmatrix} x_{vehicle} \\ y_{vehicle} \\ \theta_{vehicle} \\ x_{landmark1} \\ y_{landmark1} \\ \vdots \end{pmatrix}

My current setup allows me to use an IMU+GPS as well.

The IMU+GPS module already localizes the position of the vehicle, but I am struggling with localizing the landmarks (which I need to do in order to create a map). I would like to make use of this module that is already available to me as it enables more precise localization of the vehicle.

I was considering taking the localizations of the IMU+GPS as an input, but I had trouble coming up with the update equations, as I could not conceive of what the covariance matrix would look like.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.