Open yw9803 opened 1 week ago
@yw9803
Thanks for testing our work!
Now I've created a humble branch! :smile:
Feel free to test it out! I haven't tested ROS2 version properly, so there may be bugs.
And I have a few questions for you.
While you're experimenting, can you share the Rviz screen and what you see in the terminal when you set verbose
to true
?
If the ground segmentation works well, does it look like the height of the LiDAR from the ground was also estimated well?
@Taeyoung96 Thanks very much for your reply! I tried using your Humble version. During this process, I found two issues:
ceres::Manifold *quatManifold = new ceres::QuaternionManifold();
]Thank you so much for contributing such wonderful work and have a nice day!
@yw9803
Thanks for the detailed explanation.
It also looks like you should try to utilize gdb to find out where the error is coming from.
For example, mapping_velodyne.launch.py
,
gril_calib_node = Node(
package='gril_calib',
executable='gril_calib_exec',
parameters=[config_path,
{'use_sim_time': use_sim_time,
'point_filter_num': point_filter_num_param,
'max_iteration': max_iteration_param,
'cube_side_length': cube_side_length_param}],
prefix='gdb -ex=run --args', # Add this
output='screen'
)
@Taeyoung96 Great work! I tried to migrate your code to ROS2 and adapt it for solid-state LiDAR ([https://github.com/url-kaist/patchwork-plusplus], patchworkpp has eliminated the impact of rings, so solid-state LiDAR can also work). The LiDAR I am using is [https://www.robosense.ai/en/rslidar/E1], but I did not get good results. The configuration I used is like this. `/**: ros__parameters: common: lid_topic: "/rslidar_points" imu_topic: "/imu_data"
` Is this method not very suitable for solid-state LiDAR with a small FOV?. If you want to give it a try, here is the dataset I used https://drive.google.com/file/d/1uDVxiupJaw84qUKLMOk1S91RZ3LQOtE2/view?usp=drive_link. Thank you again for contributing such a wonderful method, and I wish you a happy life!