Systems and Methods for Clustering Objects

Patent metadata

Field	Value
Publication number	US-2025172405-A1
Application number	US-202519039884-A
Country	US
Kind code	A1
Filing date	Jan 29, 2025
Priority date	Feb 14, 2023
Publication date	May 29, 2025
Grant date	—

Abstract

Systems and methods may detect objects. In one implementation, a method may include obtaining drive information captured during drives by a plurality of vehicles traversing or having traversed a road segment. The drive information for each of the plurality of vehicles may include a drive identifier and landmark detection information corresponding to one or more landmark detections. The landmark detection information included in the obtained drive information from the plurality of vehicles may be aggregated, and based on the aggregated landmark detection information, at least two landmark clusters may be identified. The method may determine, based on the drive identifier associated with the drive information received from each of the plurality of vehicles, a distribution of drive identifiers relative to the at least two landmark cluster and determine, based on the distribution of drive identifiers, a location identifier for one or more actual landmarks positioned along the road segment.

First claim

1 . A method for determining one or more harvested landmarks, the method comprising:
obtaining drive information captured during drives by a plurality of vehicles traversing or having traversed a road segment, the drive information for each of the plurality of vehicles comprising a drive identifier and landmark detection information corresponding to one or more landmark detections; aggregating the landmark detection information included in the obtained drive information from the plurality of vehicles, and based on the aggregated landmark detection information, identifying at least two landmark clusters, each of the at least two landmark clusters being representative of a potential real world location of an actual landmark positioned along the road segment; determining, based on the drive identifier associated with the drive information received from each of the plurality of vehicles, a distribution of drive identifiers relative to the at least two landmark clusters; determining, based on the distribution of drive identifiers, a location identifier for one or more actual landmarks positioned along the road segment; storing the location identifier for the one or more actual landmarks in a map; and distributing the map to one or more autonomous vehicles for use in navigating along the road segment, the navigating comprising determining at least one navigational response based on the location identifier for the one or more actual landmarks.

2 . The method of claim 1 , wherein the at least two landmark clusters are identified based on one or more regions identified in the aggregated landmark detection information.

3 . The method of claim 2 , wherein the one or more regions are determined based on map information.

4 . The method of claim 1 , wherein the at least two landmark clusters are identified based on topology information included in the aggregated landmark detection information.

5 . The method of claim 4 , wherein the topology information is determined based on map information.

6 . The method of claim 1 , wherein the at least two landmark clusters are identified based on orientation information included in the aggregated landmark detection information.

7 . The method of claim 1 , wherein the at least two landmark clusters are identified based on a PLS analysis of the aggregated landmark detection information.

8 . The method of claim 1 , wherein the at least two landmark clusters are identified based on one or more regions identified in the aggregated landmark detection information in which an actual landmark of the one or more actual landmarks comprises features of different sizes.

9 . The method of claim 1 , wherein a count of the one or more actual landmarks positioned along the road segment for one of the at least two landmark clusters is equal to one when a same drive identifier is not included in the distribution of the drive identifiers for the one of the at least two landmark clusters.

10 . The method of claim 1 , wherein a count of the one or more actual landmarks positioned along the road segment for one of the at least two landmark clusters is greater than one when a same drive identifier is included in the distribution of the drive identifiers for the one of the at least two landmark clusters.

11 . The method of claim 1 , wherein the one or more actual landmarks include at least a traffic sign, a traffic light, a road marking, a pole, or a construction indicator.

12 . The method of claim 1 , wherein aggregating the landmark detection information included in the obtained drive information from the plurality of vehicles is based on one or more drivable paths of a road segment.

13 . The method of claim 1 , wherein the at least one navigational response includes at least one of steering, braking, or accelerating the one or more autonomous vehicles.

14 . The method of claim 1 , wherein the landmark detection information comprises one or more three-dimensional real-world coordinates corresponding to a surface of a landmark of the one or more landmarks.

15 . The method of claim 14 , wherein the one or more three-dimensional real-world coordinates determine one or more edges of the surface of the landmark.

16 . The method of claim 1 , wherein the landmark detection information comprises a type identifier identifying a type of a corresponding landmark.

17 . The method of claim 1 , wherein the landmark detection information comprises sensor data obtained by at least one sensor of one or more of the plurality of vehicles.

18 . The method of claim 17 , wherein the at least one sensor includes a camera, a radar, or a lidar.

19 . The method of claim 17 , wherein the sensor data includes at least one of images, radar data, or lidar data.

20 . A system for determining one or more harvested landmarks, the system comprising:
a memory comprising instructions; at least one processor programmed to:
obtain drive information captured during drives by a plurality of vehicles traversing or having traversed a road segment, the drive information for each of the plurality of vehicles comprising a drive identifier and landmark detection information corresponding to one or more landmark detections;
aggregate the landmark detection information included in the obtained drive information from the plurality of vehicles, and based on the aggregated landmark detection information, identifying at least two landmark clusters, each of the at least two landmark clusters being representative of a potential real world location of an actual landmark positioned along the road segment;
determine, based on the drive identifier associated with the drive information received from each of the plurality of vehicles, a distribution of drive identifiers relative to the at least two landmark clusters;
determine, based on the distribution of drive identifiers, a location identifier for one or more actual landmarks positioned along the road segment;
store the location identifier for the one or more actual landmarks in a map; and
distribute the map to one or more autonomous vehicles for use in navigating along the road segment, the navigating comprising determining at least one navigational response based on the location identifier for the one or more actual landmarks.

21 . A non-transitory computer-readable medium storing instructions that, when executed by at least one processor, cause the at least one processor to perform operations comprising:
obtaining drive information captured during drives by a plurality of vehicles traversing or having traversed a road segment, the drive information for each of the plurality of vehicles comprising a drive identifier and landmark detection information corresponding to one or more landmark detections; aggregating the landmark detection information included in the obtained drive information from the plurality of vehicles, and based on the aggregated landmark detection information, identifying at least two landmark clusters, each of the at least two landmark clusters being representative of a potential real world location of an actual landmark positioned along the road segment; determining, based on the drive identifier associated with the drive information received from each of the plurality of vehicles, a distribution of drive identifiers relative to the at least two landmark clusters; determining, based on the distribution of drive identifiers, a location identifier for one or more actual landmarks positioned along the road segment; storing the location identifi

Assignees

• Mobileye Vision Technologies Ltd

Inventors

• Eldar Avigdor
• Ben-Shachar Ido
• Shaashua Barak

Classifications

• G06V20/584

  of vehicle lights or traffic lights · CPC title

• G06V20/582

  of traffic signs · CPC title

• G01C21/3804

  Creation or updating of map data · CPC title

• G01C21/3602 Primary

  Input other than that of destination using image analysis, e.g. detection of road signs, lanes, buildings, real preceding vehicles using a camera · CPC title

• G01C21/3407 Primary

  specially adapted for specific applications · CPC title