GNSS (Video)
Data preparation,
- Download the RINEX observation <Date>.<scenario>.<place>.<receiver>.obs
- Download the Ephemeris data, e.g. SatRef from Hong Kong Lands Department
Using GNSS library for post-processing and evaluation
Visual Inertial Odometry (Video)
Data preparation,
- Download the bag file;
- Clone the VINS-Fusion repository from GitHub;
- Update camera intrinsics and extrinsics in the yaml file;
- Run the framework of VINS-Fusion:
roslaunch vins vins_rviz.launch
rosrun vins vins_node ~/catkin_ws/src/VINS-Fusion/config/vi_car/vi_car.yaml
rosbag play DATASET.bag
Also can refer to our evaluation result: https://github.com/baaixw/VINS-Fusion-for-UrbanNavDataset-Evaluation
LiDAR Inertial Odometry (Video)
-
Take scenario (a) as an example Download the rosbag UrbanNav-HK-Medium-Urban-1
-
Prepared the fine-tune LIO-SAM based on the intrinsic and extrinsic. The origianl LIO-SAM can be found at LIO-SAM
roslaunch lio_sam run_velodyne.launch
- Replay the Rosbag
rosbag play UrbanNav-HK_TST-20210517_sensors.bag
We use Chrony+GPSd to receive the NMEA+PPS signal from Ublox M8T to synchronize ROS computer with GPS time. GPS time: It consists of a count of weeks and seconds of the week since 0 hours (midnight) Sunday 6 January 1980. ROS time: unix epoch time. The Unix epoch is 00:00:00 UTC on 1 January 1970
GPS-UTC offset is 18 leap seconds for data collected in May 2021. Let's use an example to convert the GPS time to ROS time. Assuming the GPS Time is week: 2158 and seconds: 95593, the total seconds is 2158 * 604800(seconds_in_week) + 95593 = 1305253993, then we add 315964800 seconds offset between the origin of GPS time and ROS time. Finally we can get the corresponding UTC time 1621218775 (1305253993 + 315964800 -18 = 1621218775) considering 18 leap seconds.
An example of matlab code to convert gps time to utc time is provided as below,
function utctime = gps2utc(gps_week, gps_seconds)
SECONDS_IN_GPS_WEEK = 604800.0;
utctime = (gps_week * SECONDS_IN_GPS_WEEK + gps_seconds-18) + 315964800 ; % 18 leap seconds
end
If you use the raw GroundTruth data, like UrbanNav_TST_GT_raw.txt, we provide the following tools to help convert the coordinates to the ENU frame. The tool is here: UrbanNavDataset/tools/gt_vis/