Spherical Projection for Point Clouds

Spherical Projection or Front View projection is nothing but a way to
represent the 3D point cloud data into 2D image data, and so
essentially, it also acts as a dimensionality reduction method.
Spherical projection is increasingly being used in different deep
learning solutions for processing point clouds. An example of Spherical projection can be seen on the right where the cloud is shown below and the spherical image formed is shown above.

A theoretical explanation  for Spherical projection can be found on my blog here and the code can be found  on my Github handle here.