CPC G01C 21/26 (2013.01) [G05D 1/0083 (2013.01); G05D 1/0246 (2013.01); G06N 3/04 (2013.01); G06N 3/08 (2013.01); G06V 10/454 (2022.01); G06V 10/462 (2022.01); G06V 10/764 (2022.01); G06V 10/82 (2022.01); G06V 20/56 (2022.01); G06F 2218/12 (2023.01)] | 20 Claims |
1. A method comprising:
computing, using one or more neural networks and based at least on image data representative of at least a portion of an intersection in a field of view of an image sensor of a vehicle:
first data representative of first locations of first key points located on a first line segment that represents a first line feature of the intersection and second locations of second key points located on a second line segment that represents a second line feature of the intersection; and
second data representative of a first classification associated with the first line segment and a second classification associated with the second line segment;
determining, based at least on the first classification and the second classification, a first key point of the first key points on the first line segment and a second key point of the second key points on the second line segment;
generating a proposed path through the intersection based at least on connecting the first key point of the first key points to the second key point of the second key points; and
performing one or more operations by the vehicle based at least on the proposed path.
|