Sensor data based map creation and localization for autonomous systems and applications

Patent metadata

Field	Value
Publication number	US-12189018-B2
Application number	US-202217655778-A
Country	US
Kind code	B2
Filing date	Mar 21, 2022
Priority date	Mar 18, 2022
Publication date	Jan 7, 2025
Grant date	Jan 7, 2025

Abstract

One or more embodiments of the present disclosure relate to generating RADAR (RAdio Detection And Ranging) point clouds based on RADAR data obtained from one or more RADAR sensors disposed on one or more ego-machines. In these or other embodiments, the RADAR point clouds may be communicated to a distributed map system that is configured to generate map data based on the RADAR point clouds. In some embodiments of the present disclosure, certain compression operations may be performed on the RADAR point clouds to reduce the amount of data that is communicated from the ego-machines to the map system.

First claim

What is claimed is:

1. A method comprising:
generating, using one or more processors of an ego-machine, a particular RADAR point cloud based at least on RADAR data associated with a plurality of RADAR scans respectively performed by a respective RADAR sensor of one or more RADAR sensors corresponding to the ego-machine;
compressing, using the one or more processors a RADAR data packet that includes the particular RADAR point cloud;
communicating the compressed RADAR data packet to one or more devices of a map system;
decompressing, using the one or more devices of the map system, the RADAR data packet; and
generating RADAR map data based at least on the decompressed RADAR data packet and one or more other RADAR data packets, the one or more other RADAR data packets corresponding to one or more other RADAR point clouds different from the particular RADAR point cloud, the one or more other RADAR point clouds being generated using at least one of:
the one or more RADAR sensors corresponding to the ego-machine, or one or more other RADAR sensors corresponding to one or more other machines that are different from the ego-machine.

2. The method of claim 1 , further comprising determining one or more pose parameters of the ego-machine based at least on the particular RADAR point cloud.

3. The method of claim 2 , wherein the determining of the one or more pose parameters includes one or more of:
determining one or more ego-motion parameters based at least on ego-motion data associated with movement of the ego-machine, the one or more ego-motion parameters including one or more first pose parameters of the ego-machine;
determining one or more plane parameters based at least on a ground plane associated with a pose of the ego-machine, the one or more plane parameters including one or more second pose parameters of the ego-machine; or
determining one or more alignment parameters based at least on a comparison between the particular RADAR point cloud and map data associated with a geographical area, the one or more alignment parameters including one or more third pose parameters of the ego-machine.

4. The method of claim 1 , wherein the generating of the particular RADAR point cloud includes selecting RADAR data points of the RADAR data for inclusion in the particular RADAR point cloud based at least on a strength of a RADAR return signal included in the RADAR data being above a signal strength threshold.

5. The method of claim 1 , wherein the compressing of the RADAR data packet includes compressing the RADAR data packet using one or more pre-computed encoding trees.

6. The method of claim 1 , wherein the generating of the RADAR map data comprises generating the RADAR map data to include a portion of the decompressed RADAR data packet that corresponds to an object based at least on a number of RADAR data sets associated with a combination of the decompressed RADAR data packet and the one or more other RADAR data packets that indicate a presence of the object at a particular location.

7. A processor comprising:
one or more circuits to:
generate a RADAR point cloud based at least on RADAR data associated with a plurality of RADAR scans respectively performed by a respective RADAR sensor of a plurality RADAR sensors;
compress a RADAR data packet that includes the RADAR point cloud, the compressing including applying compression using one or more pre-computed encoding trees; and
communicate, from an ego-machine, the compressed RADAR data packet to one or more devices of a map system.

8. The processor of claim 7 , wherein the one or more circuits are further to determine one or more pose parameters of the ego-machine based at least on the RADAR point cloud.

9. The processor of claim 8 , wherein the determining of the one or more pose parameters includes one or more of:
determining one or more ego-motion parameters based at least on ego-motion data associated with movement of the ego-machine, the one or more ego-motion parameters including one or more first pose parameters of the ego-machine;
determining one or more plane parameters based at least on a ground plane associated with a pose of the ego-machine, the one or more plane parameters including one or more second pose parameters of the ego-machine; or
determining one or more alignment parameters based at least on a comparison between the RADAR point cloud and map data associated with a geographical area, the one or more alignment parameters including one or more third pose parameters of the ego-machine.

10. The processor of claim 9 , wherein the determining of the one or more alignment parameters includes:
obtaining a pose space, the pose space including a plurality of pose parameter sets, each respective pose parameter set including one or more hypothetical pose parameters with respect to the RADAR point cloud;
determining a cost space for the pose space, the determining of the cost space including performing a cost determination for each respective pose parameter set of the pose space, the cost determination being based at least on a comparison between the map data and the RADAR point cloud in which the RADAR point cloud is oriented based at least on the respective pose parameter set; and
determining the one or more alignment parameters based at least on the cost space.

11. The processor of claim 7 , wherein the generating of the RADAR point cloud includes selecting RADAR data points of the RADAR data for inclusion in the RADAR point cloud based at least on a strength of a RADAR return signal included in the RADAR data being above a signal strength threshold.

12. The processor of claim 7 , wherein the generating of the RADAR point cloud includes transforming the RADAR data into a common coordinate system.

13. The processor of claim 7 , wherein the generating of the RADAR point cloud includes removing one or more portions of the RADAR data that correspond to one or more dynamic objects.

14. A system comprising:
one or more processing units to:
decompress a compressed RADAR packet, the decompressing being based at least on one or more pre-computed encoding trees used to compress the RADAR data packet, the RADAR packet including one or more RADAR point clouds each associated with one or more RADAR scans respectively performed by one or more RADAR sensors; and
generate RADAR map data associated with a geographical area based at least on combined RADAR data that includes first RADAR data of the RADAR data packet and second RADAR data of one or more other RADAR data packets, the combined RADAR data including a plurality of RADAR data sets, each respective RADAR data set being obtained by one or more sensors disposed on a corresponding ego-machine during traversal through the geographical area by the corresponding ego-machine.

15. The system of claim 14 , wherein the generating of the RADAR map data includes including, in the RADAR map data, a portion of the combined RADAR data that corresponds to an object based at least on a number of RADAR data sets that each indicate presence of the object at a particular location in the geographical area.

16. The system of claim 14 , wherein the generating of the RADAR map data includes removing a portion of the combined RADAR data that corresponds to an object based at least on a number of RADAR data sets that each indicate presence of the object at a particular location in the geographical area.

17. The system of claim 14 , wherein the one or more processing units are to generate the combined RADAR data by aligning the first RADAR data and the second RADAR data.

Assignees

• Nvidia Corp

Inventors

• Akbarzadeh Amir
• Carley Andrew
• Henke Birgit
• Lu Si
• Stojanovic Ivana
• Agrawal Jugnu
• Kroepfl Michael
• Sheng Yu
• Nister David
• Zheng Enliang
• Arora Niharika

Classifications

• B60W2420/408

  Radar; Laser, e.g. lidar · CPC title

• G01S7/003

  Transmission of data between radar, sonar or lidar systems and remote stations · CPC title

• G01S2013/9316

  combined with communication equipment with other vehicles or with base stations · CPC title

• G01S13/86

  Combinations of radar systems with non-radar systems, e.g. sonar, direction finder · CPC title

• G01S7/40

  Means for monitoring or calibrating · CPC title