Is there any other way to use raw livox lidar data about fast_lio HOT 8 CLOSED

I think, if pointcloud2 contains the timestamp of every lidar point (I am not sure about that yet), you may try to use Velodyne configuration to test.

Here is the performance of the Mid70 with an external IMU without sync and calibration to the lidar, just adding the offset_time to the pointcloud2. https://www.youtube.com/watch?v=-t6ghPeJXCE

looks good👍🏻

from fast_lio.

Is there any other way to use raw livox lidar data such as /livox/lidar_xxxxxxxxxx | Type: sensor_msgs/msg/PointCloud2 directly?

We are trying now.

from fast_lio.

I think, if pointcloud2 contains the timestamp of every lidar point (I am not sure about that yet), you may try to use Velodyne configuration to test.

Here is the performance of the Mid70 with an external IMU without sync and calibration to the lidar, just adding the offset_time to the pointcloud2.
https://www.youtube.com/watch?v=-t6ghPeJXCE

from fast_lio.

Thanks very much!

from fast_lio.

I think, if pointcloud2 contains the timestamp of every lidar point (I am not sure about that yet), you may try to use Velodyne configuration to test.

from fast_lio.

Thanks for your advice, I will try to check it!

from fast_lio.

I have modified the livox_ros_driver to add the offset_time to the PointCloud2 data format, and have tested on fast_lio, thanks very much for your advices!

from fast_lio.

Hi @Tomato1107 !
I'm trying to replicate what you did. I mean, to use the pointcloud2 data format for a livox horizon capture. I have a few questions hoping you can kindly help me:

1. Regarding the modifications of livox_ros_driver to add the offset_time to the PointCloud2 data format, did you do something similar to this repo?
   I mean, did you only add the offset_time field cloud_msg.fields[0].name = "offset_time"; and then you get the offset_time similar to the function FillPointsToCustomMsg from the original repo?
2. From the FAST-LIO repo, what handler did you use? The avia_handler or the velodyne_handler as suggested by @XW-HKU
   Thank you!
   NOTE: Not sure if you have it but it would be really helpful if you already have a forked repo with the modifications you did. 🙏

from fast_lio.