ebernal-upc/lidar_l3cam_omniverse

DATASET GENERATED WITH OMNIVERSE

This dataset contains LiDAR Point Clouds and RGB images, from Beamagine's L3CAM simulated device.

The data has been created using NVIDIA Omniverse platform, with a custom extension for simulating the L3CAM.

The scene simulates a port with containers and a containers ship.

Simulated L3CAM

The L3CAM device has 2 LiDAR modules, each one with FOV of 30ºx20, total of 60ºx20º and one RGB camera with narrow FOV with focus_distance 26 and focal length 2.8

rot = rot_utils.euler_angles_to_quats(np.array([0, 0, 90]), degrees=True)

cam = Camera(frequency=15,
        resolution=(1920, 1080),
        position = (0,0,5),
        orientation = rot,
        name=name,
        prim_path=prim_path
        )
        
cam.initialize()

cam.set_focus_distance(26.0)
cam.set_focal_length(2.8)

Content

The frames have been timestamped using format yyyyMMdd_hhmmsszzz.

The timestamp is in the name of each file, in order to match images and Point Clouds, algorithms searching closest frames in time within frames can be applied.

LIDAR

Point Clouds in binary pcd format

RGB

Images in png format