Livox Mid-70 with 9axis IMU don't work

[yuataka]

Thank you for publishing this great code, but I have some trouble to run lio-sam with Livox Mid-70 and 9 axis IMU. When I ran lio-sam with Livox Mid-70 and 9axis IMU, nothing was showed on Rviz window.

First, I checked /lio_sam_imageProjection based on this issue. https://github.com/TixiaoShan/LIO-SAM/issues/110

/lio_sam/deskew/cloud_deskeded topic is not outputted from /lio_sam_imageProjection . I also checked input to /lio_sam_imageProjection . /points_raw topic is outputted, but /odometry/imu_incremental is not outputted. However, /imu is outputted correctly and inputted to /lio_sam_imuPreintegration.

Therefore, I think /lio_sam_imuPreintegration doesn't work correctly because my setting or IMU output have some problem.

Here is my config and I used your custum livox ros driver.

  # Topics
  pointCloudTopic: "points_raw"               # Point cloud data
  imuTopic: "imu"                         # IMU data
  odomTopic: "odometry/imu"                   # IMU pre-preintegration odometry, same frequency as IMU
  gpsTopic: "odometry/gpsz"                   # GPS odometry topic from navsat, see module_navsat.launch file

  # Frames
  lidarFrame: "livox_frame"
  baselinkFrame: "livox_frame"
  odometryFrame: "odom"
  mapFrame: "map"

  # GPS Settings
  useImuHeadingInitialization: false           # if using GPS data, set to "true"
  useGpsElevation: false                      # if GPS elevation is bad, set to "false"
  gpsCovThreshold: 2.0                        # m^2, threshold for using GPS data
  poseCovThreshold: 25.0                      # m^2, threshold for using GPS data
  
  # Export settings
  savePCD: false                              # https://github.com/TixiaoShan/LIO-SAM/issues/3
  savePCDDirectory: "/Downloads/LOAM/"        # in your home folder, starts and ends with "/". Warning: the code deletes "LOAM" folder then recreates it. See "mapOptimization" for implementation

  # Sensor Settings
  sensor: livox                            # lidar sensor type, 'velodyne' or 'ouster' or 'livox'
  N_SCAN: 1                                  # number of lidar channel (i.e., Velodyne/Ouster: 16, 32, 64, 128, Livox Horizon: 6)
  Horizon_SCAN: 3000                          # lidar horizontal resolution (Velodyne:1800, Ouster:512,1024,2048, Livox Horizon: 4000)
  downsampleRate: 1                           # default: 1. Downsample your data if too many points. i.e., 16 = 64 / 4, 16 = 16 / 1
  lidarMinRange: 0.3                          # default: 1.0, minimum lidar range to be used
  lidarMaxRange: 100.0                       # default: 1000.0, maximum lidar range to be used

  # IMU Settings
  imuAccNoise: 3.9939570888238808e-03
  imuGyrNoise: 1.5636343949698187e-03
  imuAccBiasN: 6.4356659353532566e-05
  imuGyrBiasN: 3.5640318696367613e-05
  imuGravity: 9.80511
  imuRPYWeight: 0.01

If there is a problem on this config, could you please tell me?

[gongyue666]

I also try to run with livox mid 70, but it seem do not support lidar data with field "ring".

[RtJinxuWang]

"ring" ：Is it the "line" of livox point cloud data?

[525753936]

support lidar data with field "ring".

So did I. Did you try to run fast-lio with livox mid 70 and IMU ? I tried, but it turned our that the odometry would have large drifts once the threads started.

[redheli]

yes, I tried run with HAP and external imu(3dm-gx5) , large drift, even when the lidar is stationary

[matiaskotlik]

@yuataka were you able to get LIO-SAM working with the Livox MID-70? Could you share your configuration and a working bag file?

Thanks.

[Griwa88]

I got Mid-70 to work on outside scan. I think the issue is all the points are quiet dense in a tight beam, this gives too many false positives and causes the drift. Also the rotating pattern isn't much help.

My parameters changes which got it to work are:

sensor: livox # lidar sensor type, 'velodyne' or 'ouster' or 'livox' N_SCAN: 1 # this is line number in custom msg, there is only 1 value '0' therefore only one 'channel' Horizon_SCAN: 9984 # MID70 9984 points per message downsampleRate: 1 # default: 1. Downsample your data if too many points. i.e., 16 = 64 / 4, 16 = 16 / 1 lidarMinRange: 0.2 # default: 1.0, minimum lidar range to be used lidarMaxRange: 250.0 # default: 1000.0, maximum lidar range to be used

edgeThreshold: 1.0 surfThreshold: 0.1 edgeFeatureMinValidNum: 1 # for MID-70 surfFeatureMinValidNum: 100

odometrySurfLeafSize: 0.2 # default: 0.4 - outdoor, 0.2 - indoor, 0.2 - outdoor MID-70 mappingCornerLeafSize: 0.2 # default: 0.2 - outdoor, 0.1 - indoor, 0.2 - outdoor MID-70 mappingSurfLeafSize: 0.2 # default: 0.4 - outdoor, 0.2 - indoor, 0.2 - outdoor MID-70

Also, I wrote a Subscriber/Publisher node to take Livox Custom Message and translate it to PointXYZIRT, but you can use Tixiao custom driver: https://github.com/TixiaoShan/livox_ros_driver [I'm not publishing my Sub/Pub node, maybe in the future once I clean it up, as it was my first go at C++, but the idea is I can use custom message for other localizations like FastLio and LioSam without having to rescan.] x->x y->y z->z intensity->intensity line->ring time / 1000000000->time

Hope this helps.

[yuataka]

Sorry, currently I am not trying to run lio-sam with Livox Mid-70. So, I close this issue. Thank you.