hyye / lio-mapping

Implementation of Tightly Coupled 3D Lidar Inertial Odometry and Mapping (LIO-mapping)
https://sites.google.com/view/lio-mapping
GNU General Public License v3.0
897 stars 320 forks source link

Question about point mapping #64

Closed DarrenWong closed 4 years ago

DarrenWong commented 4 years ago

Hey All,

I am learning the part of point_processor together with the LOAM paper https://www.ri.cmu.edu/wp-content/uploads/2019/01/AURO_2017_2.pdf

In the below variables I might a little confused and hope someone can give me a hand.

https://github.com/hyye/lio-mapping/blob/8485a4d9c359b7e486ad019a10815ecb8fd1257f/src/point_processor/PointMapping.cc#L77-L82

As for my understanding, the laser_cloudlength, laser_cloudwidth and laser_cloudheight consist of the number of cubes in the existing cloud on the map, and the laser_cloud_cenlength, laser_cloud_cenwidth and laser_cloud_cenheight is the center of the existing cloud on the map when init. (unit in meters) Attached my view of the meaning of the variables, front_view top_view

My question is, In the paper (part 6 Lidar mapping, as attached pic), it would be a 10cm 10cm 10cm certain region to find the cloud then insert to KD-tree. Some explanation say that it would be corresponding to the laser_cloud_cenlength, laser_cloud_cenwidth and laser_cloud_cenheight . If that correct? It seems 10cm 10cm 10cm would be very small for insert to KD-tree. If not, where can I find the related code with the paper's 10cm 10cm 10cm certain region?

image