CPC G06N 3/08 (2013.01) [G05D 1/0221 (2013.01); G05D 1/0276 (2013.01); B60W 60/001 (2020.02); B60W 2556/45 (2020.02); G06F 7/544 (2013.01)] | 15 Claims |
1. A system for operating an autonomous platform in an environment using information extracted from point cloud data in the environment, the system comprising:
a sensor generating point cloud data; and
one or more processors and a non-transitory computer-readable medium having executable instructions encoded thereon such that when executed, the one or more processors perform operations of:
training a neural network model with imageless data, the imageless data being exclusively the point cloud data, comprising:
using a parallelizable k-nearest neighbor sorting algorithm, providing a patch of points sampled from the point cloud data as input to the neural network model, wherein the patch of points has a central point;
transforming the patch of points from Euclidean coordinates in a Euclidean space to spherical coordinates; and
generating normalized points by projection onto the unit sphere;
estimating a polar angle of a surface normal of the normalized points in the spherical coordinates;
utilizing the trained neural network model on the autonomous platform; and
using the estimate of the polar angle of the surface normal to guide operation of the autonomous platform within the environment.
|