0
$\begingroup$

Rosanswers logo

I want to have hector slam loaded with a pre-saved map , but NOT for localization only (so can't go to AMCL) , it is for performing mapping and update the saved map with the new info found in the world that weren't in the saved map (the world changed) .

To be specific : suppose a robot scanned the map of an entire floor with hector slam and saved the map successfully . Now next time the robot will use the saved map , BUT the floor itself changed a little bit (positions of obstacles changed, new obstacles or new free areas rise) so I want be aware and map those to the saved map , sort of logical OR operator between the saved map and hector's incremental findings.

If I tried just publish the saved map to the hector's map topic , hector will overwrite it with its own map , so maybe if there's a way to just make hector take the last msg from the map topic , (logical or ) to add the new updates , and then publish it .

This will not make it a closed loop in the sense that it closed the loop to improve performance in any way , its just publishing updates on the saved thing instead of publishing a new thing


Originally posted by amhadR on ROS Answers with karma: 3 on 2020-04-17

Post score: 0

$\endgroup$

1 Answer 1

0
$\begingroup$

Rosanswers logo

You can look in to Jin-W-FS's Pull Request at https://github.com/tu-darmstadt-ros-pkg/hector_slam/pull/45

I am also currently trying to implement the features that you have mentioned.


Originally posted by jjbecomespheh with karma: 18 on 2020-07-06

This answer was ACCEPTED on the original site

Post score: 0


Original comments

Comment by amhadR on 2020-08-04:
great ! I'll mark this as an answer since I didn't find more useful resources.

It seems that the PR works well when the robot starts from the same initial pose of the saved map , but I still have to try myself , did you try it and can confirm?

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.