Help needed - How to enable taking further feature points into computation

at-wat / mcl_3dl

A ROS node to perform a probabilistic 3-D/6-DOF localization system for mobile robots with 3-D LIDAR(s). It implements pointcloud based Monte Carlo localization that uses a reference pointcloud as a map.

BSD 3-Clause "New" or "Revised" License

486 stars 119 forks source link

Help needed - How to enable taking further feature points into computation #323

Open HappySamuel opened 4 years ago

HappySamuel commented 4 years ago

When my robot is far away from the building features, the scan-to-map matching become less depend on them, which give a bad localization result. Is there a parameter to adjust for that? Or how shall i amend the code to achieve that?

Best, Samuel

at-wat commented 4 years ago

https://github.com/at-wat/mcl_3dl/blob/b192ce82cee92ff00d21c39e432bc1d560ea2c3a/src/parameters.cpp#L58-L74

likelihood/clip_far and beam/clip_far should be increased to use the points far from the robot. likelihood/match_dist_min may also be needed to be increased for large environment. Points which are not within likelihood/match_dist_min from the map will be ignored in the likelihood calculation.

HappySamuel commented 4 years ago

Hi @at-wat

For example, for likelihood/clip_far as 50m, how much shall i set for likelihood/match_dist_min ? for likelihood/clip_far as 100m, how much shall i set for likelihood/match_dist_min ?