Hello,

I am currently trying to evaluate different SLAM algorithms (Gmapping, Karto-SLAM etc.).

I have used scan matching with NDT methods to estimate the robot's pose(see image), and then try to match the two scans so that I can calculate the robot's pose error using the Hausdorff distance or another method.

is this method correct for evaluating SLAM algorithm on localization aspect?

if not, what do i need to change in the methodology to evaluate SLAM on localization part?

I hope you can help me

Thank you