IP Library Granted Patent US 12,339,135
Granted Patent B2
US 12,339,135 · App. 17/823,008 · Granted Jun 24, 2025

Navigable boundary generation for autonomous vehicles

Inventors: Derek Thomas Miller (Palo Alto, CA); Lin Yang (San Carlos, CA); Mark Damon Wheeler (Saratoga, CA)
Assignee: NVIDIA CORPORATION
G01C21/3815G01C21/3848G01C21/3874G01C21/3881G05D1/0088G05D1/0214G08G1/096805G08G1/0129
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,339,135
App. No.
17/823,008
Granted
Jun 24, 2025
Kind
B2
Abstract

A system accesses a three-dimensional map of a geographic region and generates a two-dimensional projection of the road based on the three-dimensional map. The two-dimensional projection comprises a plurality of points along the road and each point is assigned a score measuring a navigability of the point. Based on the navigability score of each point and history of vehicle positions on the road, the system identifies a plurality of navigable points on the two-dimensional projection of the road. Based on the plurality of navigable points, the system determines a navigable surface corresponding to a physical area over which a vehicle may safely navigate and navigable surface boundaries surrounding that area. The navigable surface area and boundaries on the two-dimensional projection are converted into a three-dimensional representation, which the system uses to generate an updated three-dimensional map of the geographic region.

Claims (34)

1. A method comprising:

performing, by a machine, one or more navigation, localization, or control operations for maneuvering the machine based at least on map data of an area, wherein the map data is updated at least by:

determining a navigability score corresponding to one or more locations based at least on a relationship between a first orientation characteristic of the one or more locations and a second orientation characteristic of a ground plane;

determining that the one or more locations correspond to a navigable surface based at least on the navigability score; and

updating the map data with respect to the one or more locations based at least on the determining that the one or more locations correspond to the navigable surface.

2. The method of claim 1 , wherein the first orientation characteristic includes a first normal corresponding to the one or more locations and the second orientation characteristic includes a second normal corresponding to the ground plane.

3. The method of claim 2 , wherein the relationship between the first orientation characteristic and the second orientation characteristic includes an angular deviation between the first normal and the second normal.

4. The method of claim 1 , wherein:

the first orientation characteristic is determined based at least on a first point included in a point cloud that corresponds to a region including the one or more locations; and

the second orientation characteristic is determined based at least on a second point included in the point cloud and that corresponds to the ground plane within at least a portion of the region.

5. The method of claim 4 , further comprising using the first point as a reference point for identifying the navigable surface based at least on the first point being identified, based at least on the navigability score, as a navigable seed point.

6. The method of claim 5 , wherein the first point is identified as the navigable seed point based at least on one or more of:

the navigability score satisfying a threshold; or

an indication that one or more machines have previously traversed an area corresponding to the first point.

7. One or more processing units to perform operations comprising:

determining a navigability score corresponding to one or more locations based at least on a relationship between a first characteristic of the one or more locations and a second characteristic of a ground plane, the first characteristic and the second characteristic being determined based at least on a point cloud that corresponds to a region including the one or more locations and the ground plane; and

causing a machine to perform one or more maneuvering operations with respect to the one or more locations based at least on the navigability score.

8. The one or more processing units of claim 7 , wherein the performing of the one or more operations with respect to the one or more locations includes one or more of:

updating map data corresponding to the one or more locations with information indicating that the one or more locations correspond to a navigable surface.

9. The one or more processing units of claim 7 , wherein the performing of the one or more operations with respect to the one or more locations is based at least on a determination that the one or more locations correspond to a navigable surface, the determination that the one or more locations correspond to the navigable surface being based at least on the navigability score.

10. The one or more processing units of claim 7 , wherein the first characteristic includes a first normal corresponding to the one or more locations and the second characteristic includes a second normal corresponding to the ground plane.

11. The one or more processing units of claim 7 , wherein the operations further comprise using a point included in the point cloud as a reference point for identifying a navigable surface based at least on the point being identified, based at least on the navigability score, as a navigable seed point.

12. The one or more processing units of claim 11 , wherein the first point is identified as the navigable seed point based at least on the navigability score satisfying a threshold.

13. The one or more processing units of claim 11 , wherein the first point is identified as the navigable seed point based at least on an indication that one or more machines have previously traversed an area corresponding to the first point.

14. A system comprising:

one or more processors to perform operations comprising:

determining a navigability score corresponding to one or more locations based at least on a relationship between a first orientation characteristic of the one or more locations and a second orientation characteristic of a ground plane; and

updating map data that corresponds to the one or more locations based at least on the navigability score, wherein a machine performs one or more navigation, localization, or control operations for maneuvering based at least on the map data as updated.

15. The system of claim 14 , wherein the updating of the map data is based at least on a determination that the one or more locations correspond to a navigable surface, the determination that the one or more locations correspond to the navigable surface being based at least on the navigability score.

16. The system of claim 14 , wherein the first orientation characteristic includes a first normal corresponding to the one or more locations and the second orientation characteristic includes a second normal corresponding to the ground plane.

17. The system of claim 14 , wherein:

the first orientation characteristic is determined based at least on a first point included in a point cloud that corresponds to a region including the one or more locations; and

the orientation second characteristic is determined based at least on a second point included in the point cloud and that corresponds to the ground plane.

18. The system of claim 14 , wherein the operations further comprise using a point as a reference point for identifying a navigable surface based at least on the navigability score.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 27, 2025
From: MILLER, DEREK THOMAS; YANG, LIN; WHEELER, MARK
To: DEEPMAP INC.
Reel/Frame 072692/0570 →
Continuity (3)
Continuation 16721516 · Dec 19, 2019
Provisional Application 62782331 · Dec 19, 2018
Related Publication 20220412747A1 · Dec 29, 2022
References Cited (6)
US 10474164B2 · Wheeler · 2019 [cited by applicant]
US 20140297116A1 · Anderson et al. · 2014 [cited by applicant]
US 20160054135A1 · Fowe · 2016 [cited by examiner]
US 20170146360A1 · Averbuch · 2017 [cited by applicant]
US 20180188742A1 · Wheeler · 2018 [cited by applicant]
US 20230196148A1 · Fields · 2023 [cited by examiner]