US 11,754,719 B2
Object detection based on three-dimensional distance measurement sensor point cloud data
Sandeep Gangundi, Milpitas, CA (US); Sarthak Sahu, Palo Alto, CA (US); Nathan Harada, San Francisco, CA (US); and Phil Ferriere, Palo Alto, CA (US)
Assigned to GM Cruise Holdings LLC, San Francisco, CA (US)
Filed by GM Cruise Holdings LLC, San Francisco, CA (US)
Filed on Jul. 30, 2021, as Appl. No. 17/389,907.
Application 17/389,907 is a continuation of application No. 16/795,129, filed on Feb. 19, 2020, granted, now 11,107,227.
Prior Publication US 2021/0358145 A1, Nov. 18, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G01S 17/894 (2020.01); G06T 7/50 (2017.01); G06T 15/08 (2011.01); G06V 20/58 (2022.01); G06V 10/80 (2022.01); G06V 10/82 (2022.01); G06V 10/44 (2022.01); G01S 15/89 (2006.01); G01S 13/89 (2006.01); G01S 17/89 (2020.01)
CPC G01S 17/894 (2020.01) [G06T 7/50 (2017.01); G06T 15/08 (2013.01); G06V 10/454 (2022.01); G06V 10/803 (2022.01); G06V 10/82 (2022.01); G06V 20/58 (2022.01); G01S 13/89 (2013.01); G01S 15/89 (2013.01); G01S 17/89 (2013.01); G06T 2200/04 (2013.01); G06T 2207/10028 (2013.01); G06T 2207/30261 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method of object detection, the method comprising:
receiving a first plurality of distance measurements from a first distance measurement sensor of a plurality of distance measurement sensors;
generating a first three-dimensional point cloud including a first plurality of points in a volume of space around the plurality of distance measurement sensors, wherein each of the first plurality of points in the first three-dimensional point cloud corresponds to one of the first plurality of distance measurements;
receiving a second plurality of distance measurements from a second distance measurement sensor of the plurality of distance measurement sensors;
generating a second three-dimensional point cloud including a second plurality of points in the volume of space around the plurality of distance measurement sensors, wherein each of the second plurality of points in the second three-dimensional point cloud corresponds to one of the second plurality of distance measurements;
generating a voxelized model of the volume of space based on a combination of the first three-dimensional point cloud and the second three-dimensional point cloud, the voxelized model including a plurality of voxels positioned based on the first plurality of points and the second plurality of points; and
identifying an object within the voxelized model by classifying a cluster of voxels using a model, wherein at least a portion of the voxelized model, which is missing one or more points from at least the first three-dimensional point cloud or the second three-dimensional point cloud is filled in with voxels based on the classification of the cluster of voxels.