MS-Dataset

January 22, 2025 ยท View on GitHub

Large scale LiDAR-IMU dataset

image (15)

sensor configuration

  • Pandar XT32, +-0.5cm
  • SBG INS : 100Hz IMU + GNSS + EKF filter
  • 3DM-GQ7 INS: 700Hz IMU + dual GNSS-RTK + 30Hz Filter, >= 1.4cm
  • Mono Flir RGB camera (optional), it's hard to collect data with RTK-GNSS and RGB camera at the same time due to USB3.0.
  • Ground Truth RGB Point Cloud Map (indoor -> outdoor, 4mm), collected by Leica BLK360 and RTC360.

GT Trajectory

  • From INS with cov less than 0.005m. Users can also generate the gt with my scripts.
  • Users can generate GT either from RTK-GNSS 1 or RTK-GNSS 2 with my scripts, but only with x y z yaw .
  • KML file which can directly import by google map with my scripts.
3DM-INSGNSS1GNSS2SBG-GNSS
RB02image-20250122202712468
RB03image-20240624131357960image-20240624131801414image-20240624131343201image-20240624131412226
PK01image-20240627071947142image-20240627072000320
CP05image-20240627072713997image-20240627072622052image-20240627072635693image-20240627072730361
CP02image-20240627072854370image-20240627072809607image-20240627072821714image-20240627072834323
RD01image-20240627073000945image-20240627072926561image-20240627072940627
CP01image-20240627073107324image-20240627073035768image-20240627073047934
GC01image-20240627073207972image-20240627073136950image-20240627073146802image-20240627073157026
CC01image-20240627073243335

Data Description

DatasetDescriptionSensorsDownload LinksGround TruthComments
Calib-IMU-Intrinsinc
Calib-Camera-LiDAR
PK01Degenerate parkinglot
RB02
RB03
CP05

Issues

How to adpate to different sensors

rostopic echo /os1_cloud_node/points/fields -n1