I've had a look at the laser_line_extraction package and it extracts line segments from a single laser scan message. How are you building a map out of these lines?
There seems to be a step missing. The gmapping package aligns many laser scans with each other to create a single global map. It uses a regular grid to store this map so many scans can be easily added to it.
To make a vector based map you'd have to do the same steps but with line segments which is far more complicated to implement.
To save these maps you would need to use a standard vector format (svg, dxf, etc) or your own custom one. The pgm file format used by gmapping stores raster images so cannot store vector information.
I don't know of any out of the box ROS system that can achieve what you want but it would be a good project to get stuck in and build them.
Hope this makes sense.
How are you storing these maps in your program as they are built? Are they just a list of line segments or do you have additional meta data such as confidence or age?
I only have the following topics that the /line_extractor publishes : * /line_markers [visualization_msgs/Marker] * /line_segments [laser_line_extraction/LineSegmentList] Is there any way to save these line segments into a yaml or pgm file like you do with the gmapping package?