Semantic_point_annotator Parameter Explanation

edit

I've been having some problems with the Semantic Point Annotator, i.e. the thing that removes the ground plane from the PR2's obstacle avoidance system. The main problem is that points in the floor show up as obstacles.

The wiki page does not have much in the way of documentation. Does anyone have insight into how the parameters can be fiddled with to remove this problem?

Edit: Alternatively, you can check out the bag file