## Octree-based lossy point-cloud compression with open3d and numpy

Average compressing rate (depends on octreee depth setting parameter): 0.012 - 0.1   

(Original article: Szutor Péter: Ipari pontfelhők tömörítése,  
Dr. Balázs Boglárka (szerk.) Az elmélet és a gyakorlat találkozása a térinformatikában VIII. Theory meets practice in GIS Debrecen, Magyarország  
Debreceni Egyetemi Kiadó, (2017) pp. 349-355)

Input formats: You can get a list of supported formats from : http://www.open3d.org/docs/release/tutorial/Basic/file_io.html#point-cloud
               (xyz,pts,ply,pcd)

Usage:
    
Dependencies: Open3D, Numpy  (You can install theese modules:  pip install open3d, pip install numpy)    
    
Compress a point cloud:
    
octreezip(<filename>,<depth>) -> <result>
<filename>: (str) Point Cloud file name. Saved file name: [filename without ext]_ocz.npz  (Yes, it's a numpy array file)
<depth>   : (str) Octree depth. You can try 11-16 for best result. Bigger depht results higher precision and bigger compressed file size.
<result>  : (str) If the compressing was success you get: "Compressed into:[comp.file name] | Storing resolution:0.003445". Storing resolution means the precision.
                  The PC file is missing or bad: "PC is empty, bad, or missing"
                  Other error: "Error: [error message]"
                  

Uncompressing:
octreeunzip(<filename>) -> <result>
<filename>: (str) Zipped Point Cloud file name (npz). Saved file name: [filename].xyz  (standard XYZ text file)
<result>  : (str) If the compressing was success you get: "Saved: [filename].xyz"
                  Other error: "Error: [error message]"