Question about generating KITTI range view images

[tengteng95]

Thanks for your great work. Could you please tell me the meaning of height and incl in the function get_range_image [code] ?

[Abyssaledge]

height means the vertical height of each laser beam.
incl means the inclination of each laser beam.
Note that the 64 laser beams of LiDAR are not emitted from a single point.

[Devoe-97]

height means the vertical height of each laser beam. incl means the inclination of each laser beam. Note that the 64 laser beams of LiDAR are not emitted from a single point.

Thanks for your great work. Could you please tell me how to calculate the height and zenith? Can the nuscenes data be converted into range view data format? What should I do for it?

[Abyssaledge]

@HuangDihe The heights and zeniths we used in KITTI are provided by the authors of RCD, which is obtained via Hough Transformation. You need to estimate height and zenith for each beam, which is somewhat similar to line detection with Hough Transformation. I believe you can do it in NuScenes but I cannot provide you the source code.

[qrzeffort]

@Abyssaledge Thanks for your great work! I run the code ：
python ./datasets/create_range_image_in_kitti.py --source-dir data/kitti --target-dir output
but there is no results in --target-dir(only have npz_test,npz_trainval folder with no files and have .roidb files（5 Byte）)，and report：

Error: 'lidar_idx'

the kitti data organized as follow,and generated .pkl files
data
│ ├── kitti
│ │ │── ImageSets
│ │ │── training
│ │ │ ├──calib & velodyne & label_2 & image_2 & (optional: planes) & (optional: depth_2)
│ │ │── testing
│ │ │ ├──calib & velodyne & image_2

[QYChan]

@Abyssaledge What is the version of mmdet3d? The kitti pkl file changed in different version.

[Petros626]

@Devoe-97 did you succeed?
@Abyssaledge can you tell me how to convert KITTI .bin pc to a range image, I always get a translated result with def get_range_image(pc, incl, height):.

Original Point Cloud:

Reconstructed Point Cloud: