[1] Platform


Hyndai, IONIQ 2015 electric model is used for data acquisition.



[2] Sensor Configuration

Hardware sensor configuration of MSC-RAD4R dataset with top view. Sensor coordinates of each sensor are represented, where the coordinate frame is shown with red(x), green(y) and blue(z) axes. The colored triangles represent the field of view of each sensor, yellow for camera and red for 4D Radar.





[3] Sensor Specifications

Sensors Specifications Hz Topic Name Message Type N
Camera FLIR, BFS-16S2C, HFoV, VFoV 60x45, resolutions 720x540 15 /camera_array/(left, right)/image_raw /sensor_msgs/Image 2
LiDAR Ouster, OS1-128, VFoV 45, range 200m 10 /ouster/points /sensor_msgs/PointCloud2 1
4D Radar Oculii, Eagle, HFoV, VFoV 113x45, long range 400m 15 /oculii_radar/point_cloud /sensor_msgs/PointCloud2 1
IMU Xsens, MTi-630, roll(0.2), pitch(0.5), yaw(1) 200 /imu/data /sensor_msgs/Imu 1
RTK-GPS Synerex, MRP-2000, RTK accuracy +- 3cm 10 /fix, /vel /sensor_msgs/NavSatFix, /geometry_msgs/TwistStamped 1
GPS UBlox, ZED-F9P, Pos. accuaracy 1.0 - 1.5m 10 /ublox_gps/fix, /ublox_gps/fix_velocity /sensor_msgs/NavSatFix, /geometry_msgs/TwitWithCovarianceStamped 1
Wheel Speed Autonics, PR30-10DN, detection dis. 10mm 100 /wheel_(left,right) /geometry_msgs/TwistStamped 2





[4] 4D Radar Data Format

Originally x, y and z are measured in Oculii coordinate, but these values were changed to LiDAR coordinates for the convenience of the ROS visualizer.

name description units
x oculii_x [m]
y oculii_y [m]
z oculii_z [m]
beta elevation angle [degree]
alpha azimuth angle [degree]
power signal-to-noise(SNR) [dB]
doppler doppler velocity in radial direction [m/s]
range distance of point [m]
recoveredSpeed recovered(absolute) speed in radial direction [m/s]
denoiseFlag flag for moving objects or noise 1 or 0