search

RobustFieldAutonomyLab / LeGO-LOAM

LeGO-LOAM: Lightweight and Ground-Optimized Lidar Odometry and Mapping on Variable Terrain
BSD 3-Clause "New" or "Revised" License
2.35k stars 1.11k forks source link

The point clouds are vertical after records as pcd files. #139

Closed xingtongli closed 4 years ago

xingtongli commented 4 years ago

HI, I am working with Autoware and I want to use this algorithm to prepare the pcd map for using. I met a problem. I tried to use pointcloud_to_pcd to record and export the pointclouds as pcd files. After that, I import the pcd file into autoware. However I found that the pointclouds map is vertical. image When I runing the algorithm, I saw the pointclouds are horizental in Rviz. Do you rotate the point clouds in the algorithm? I recorded the pointclouds from topic: laser_cloud_surround. What is the reason of this situation?

TixiaoShan commented 4 years ago

Yes, the points are rotated here https://github.com/RobustFieldAutonomyLab/LeGO-LOAM/blob/744254789c91ebe4851b4678be2c9f97e2f56c27/LeGO-LOAM/src/featureAssociation.cpp#L487.

xingtongli commented 4 years ago

Thank you for your reply, can you give me some hint of how to modify it? It does not work if I only modify the x,y,z.

TixiaoShan commented 4 years ago

@xingtongli You need to re-write all the code that is related to point cloud computation. It's probably not that easy.