Draco Data Compression on Lidar


I am trying to run Draco ( link : https://github.com/google/draco ) compression algorithm on Lidar part of Nuscenes dataset. However the lidar files are provided is pcd.bin file format.

Is there any way to convert .pcd.bin files to ply file format (Polygon File Format) ? As Draco works with ply file format.


Hi, we don’t provide a converter, but our data is really just a numpy error, which we found out is the most efficient format for loading. You can load the lidar pointclouds with

        scan = np.fromfile(file_name, dtype=np.float32)
        points = scan.reshape((-1, 5))[:, :4]
        return points.T

Then you can write your own tool to export to ply.

Thank you for the reply.

I understand that pcd.bin file is created by removing header information from .pcd lidar files. Can you please let me know where can I find header information like how many fields are present in this file.

Thanks in advance.

Hi, please take a look at the code that loads them:

“Loads LIDAR data from binary numpy format. Data is stored as (x, y, z, intensity, ring index).”

Thanks for the immediate response.

Are there standard library to convert pcd to ply format ? Also what are standard LIDAR data formats ?


PCD is a standard format, so there should be a number of libraries. I suggest to google it. I think it is good enough for lidar. The reason we used the binary numpy format was that it is the fastest to load in Python.