I use this Costmap to generate placement poses for my complete pick-and-place demo on the real robot. The outer edges are reduced because the robot requires some clearance, especially when dealing with structures like shelves where the outer areas might be restricted. Since the standard semantic costmap is no longer in use, perhaps we should consider renaming this to something like placeSemanticCostmap, @Tigul. In the attached image, you can see an example for a countertop. Please note that I work with TF frames oriented according to the direction we agreed upon within the Suturo project.
I use this Costmap to generate placement poses for my complete pick-and-place demo on the real robot. The outer edges are reduced because the robot requires some clearance, especially when dealing with structures like shelves where the outer areas might be restricted. Since the standard semantic costmap is no longer in use, perhaps we should consider renaming this to something like placeSemanticCostmap, @Tigul. In the attached image, you can see an example for a countertop. Please note that I work with TF frames oriented according to the direction we agreed upon within the Suturo project.