search

HKUST-Aerial-Robotics / Fast-Planner

A Robust and Efficient Trajectory Planner for Quadrotors
GNU General Public License v3.0
2.4k stars 665 forks source link

Clarifications regarding plan_env #67

Closed FaboNo closed 3 years ago

FaboNo commented 3 years ago

I am looking at sdf_map.cpp and I have few questions

  1. Concerning the topic /sdf_map/occupancy which is advertising pointCloud2, from what I can see the point cloud are related to the inflated map already - see screenshot below in which I superimposed the depth map and the "occupancy map" occupancy

  2. So what is the difference between /sdf_map/occupancy and /sdf_map/occupancy_inflate? The only difference I can see is the parameter local_map_margin which is divided by 2 in one case. Is it related to the fact that you can display a map in a greater area with /sdf_map/occupancy_inflate? Besides what is the meaning of local_map_margin?

  3. The parameter obstacle_inflation should be at least equal the radius of the circle centered on the drone? Am I right because in the simulation this parameter is really small (0.09).

4.I am a bit confused about md_.local_boundmin and md_.local_boundmax. I can see that they contain the indices of the position of the (min, max) of a cube but how is it related to the position of the drone?

Thanks you for your support and for this amazing piece of work

ZbyLGsc commented 3 years ago

occupancy map inflated by "obstacle_inflation" produces occupancy_inflate. local_boundmin/max_ are related to the updated regions of each frame.