• "root_path": where to save the generated data
• "sensor_names": which sensors you want to use for collecting data

• vehicle_id...
  • camera
    • image...
    • image_meta...
  • camera_sem
    • images_sem_label...
    • image_sem_meta...
  • lidar
    • pointcloud...
    • pointcloud_meta...
  • lidar_sem
    • pointclouds_sem_label...
    • pointclouds_sem_meta...
• info.csv

• line1: frame, timestamp, fov, height, width
• line2: roll, pitch, yaw, x, y, z

• line1: frame, timestamp, horizontal_angle, n_channels
• line2: roll, pitch, yaw, x, y, z
• line3: 64 elements array, each element indicate the points number measured by the corresponding laser.

the formatted data has 3 folders

• cloud_ego: clouds collected by ego vehicle, no down-sampling
• cloud_coop: point clouds of cooperative vehicles
• cloud_fused: fused clouds(including ego-vehicle cloud), down-sampled through voxel grid(size=0.2m)
• label_box: ground truth bounding boxes of all vehicles in the scene of each frame

point cloud files:

• all data are written in binary files. Data are all in float32 length,
• each data point has 4 columns(x,y,z,label), following is the semantic meaning of labels

    LABEL_COLORS = np.array([
    (0, 0, 0),       #0 Unlabeled
    (70, 70, 70),    #1 Buildings
    (100, 40, 40),   #2 Fences
    (55, 90, 80),    #3 Other
    (220, 20, 60),   #4 Pedestrians
    (153, 153, 153), #5 Poles
    (157, 234, 50),  #6 RoadLines
    (128, 64, 128),  #7 Roads
    (244, 35, 232),  #8 Sidewalks
    (107, 142, 35),  #9 Vegetation
    (0, 0, 142),     #10 Vehicles
    (102, 102, 156), #11 Walls
    (220, 220, 0),   #12 TrafficSign
    (70, 130, 180),  #13 Sky
    (81, 0, 81),     #14 Ground
    (150, 100, 100), #15 Bridge
    (230, 150, 140), #16 Railtrack
    (180, 165, 180), #17 GuardRail
    (250, 170, 30),  #18 TrafficLight
    (110, 190, 160), #19 Static
    (170, 120, 50),  #20 Dynamic
    (45, 60, 150),   #21 Water
    (145, 170, 100)  #22 Terrain
    ])

filename: <junction_id>_<frame_id>.<extention>

label_box format:

• each row indicates one bounding box,
• column from left to right are: vehicle_id, vehicle_class, x, y, z, rx, ry, rz, l, w, h
• x,y,z: the center point of the bounding box
• rx, ry, rz: orientation
• l, w, h: vehicle size