2D PointCloud to Occupancy grid

Question

Hi there,

I'm having trouble finding a package that does what I want. I'm using the Intel's Euclid to scan the room and I extract every possible plane with help of the PointCloud Library. I can store those planes for later inspection. What I need to do now is creating an occupancy grid, i.e. a matrix with cells that represent 1cm^2 with the probability of occupancy (0-100) I researched and found the octomap_server and the ethz-asl/grid_map package but I'm not sure they serve my purpose.

Any ideas? Thanks in advance

---

Originally posted by Georgee on ROS Answers with karma: 25 on 2018-04-09

Post score: 0

Answer 1

You can use Gmapping, It performs SLAM with odometry and 2D laser scanner. It gives the 2D map and entropy as well. You can change the parameters inside the programs or in .yml file.

---

Originally posted by Gayan Brahmanage with karma: 929 on 2018-04-09

This answer was ACCEPTED on the original site

Post score: 0

---

Original comments

Comment by Georgee on 2018-04-09:
Inputs for Gmapping are tf (tf/tfMessage) and scan (sensor_msgs/LaserScan), and I would like to use my 2D PointCloud plane. Any alternatives?

Thanks for your help.

Comment by gvdhoorn on 2018-04-09:
Not pushing you to use gmapping, but there is something called pointcloud_to_laserscan.

Comment by Georgee on 2018-04-10:
Thanks for your answer @gvdhoorn , but it doesn't seem to match what I'm looking for. What I have is a dataset of points that represent a plane (for indoor purposes, to be more precise, e.g a table or a wall) and I need to know how probable it is for the robot to wander through that plane.

Comment by gvdhoorn on 2018-04-10:
My comment was in response to your "I have a pointcloud, not a laser scanner" comment earlier.