Options for saving map or saving the path

[robertsenps]

Hi developers,

First of all, I'm happy to find a working 3d navigation package on the Internet after some time searching with no result. I'm excited to get this package that can check each pointclouds whether it's an obstacle or it's a traversable area. The tm_planner is also working with the package and I can navigate through the explored path.

For navigation thing, I think it's better to combine online and offline method (online: to mark dynamic obstacles, and offline: to make the robot creates a plan in a right way, because if we plan through the unexplored area, sometimes it will fail to create a right path).

I also saw the latest issues issued by Yeah2333 about saving the map and saving the prm graph. Could you explain more about how to save the map or the graph, or probably you can share the prototype code for saving the map?

Any help from you will be very helpful.

Thank you.

[Yeah2333]

Hi @robertsenps ， if you have question about saving the map and saving the prm graph. you can email me (wangzy@stu.xmu.edu.cn) or just @Yeah2333 in issue. You can explore the issue about saving the map and saving the prm graph that i have chat with the repo owner. I use rosbag to save this two things, so i have writer many custome rosmsgs. If you have any question , just email or reply me.

[rock19970106]

Hi, @Yeah2333 , How do you deal with dynamic obstacles ？ And , only by simple threshold judgment, I feel that the effect of road edge detection is not so good . Do you have the same problem ？

[Yeah2333]

@rock19970106 HI，i don't do anything about dynamic obstacles. But i have record data when the area is clear, so there is no dynamic obstacles. And when driving robot, the avoiding obstacles method in traversability_mapping using local 2d grid map is enough to deal with dynamic obstacles

[rock19970106]

Hi, @Yeah2333 , could you share the code to me too ? My email is zhlupeng@mail.ustc.edu.cn. Thank you very much.

[bimalka98]

Function for saving the PRM graph after the pilot run: void savePRMGraph()
Function for loading it in subsequent autonomous navigation sessions: void loadPRMGraph()
Source code; Tested with an actual robot: See the demo