Unified radar perception architecture

Patent metadata

Field	Value
Publication number	US-12153121-B2
Application number	US-202217591335-A
Country	US
Kind code	B2
Filing date	Feb 2, 2022
Priority date	Feb 2, 2022
Publication date	Nov 26, 2024
Grant date	Nov 26, 2024

Abstract

The present disclosure is directed to combining the strengths of different methods of analyzing collected sensor data to updating a driving pattern of an automated vehicle (AV). This may include combining data from sets of data that track movement of objects over time with instantaneously received sensor data based on a series of steps that include accessing data that tracks the motion of objects in the field of view of a sensing apparatus, receiving current sensor data that includes a component of current or instantaneous object motion, and identifying whether controls of AV should be maintained or changed. When controls of the AV are maintained, an AV may be controlled to stay in driving in a same lane of a roadway at a same velocity. When controls of an AV are changed, changes may include applying, increasing a velocity, or altering the course of the AV.

First claim

What is claimed is:

1. A method for improving perception of a sensing apparatus, the method comprising:
receiving radar sensor data associated with the current scene;
generating enriched data frame data of the current scene from the received radar sensor data;
accessing tracking data associated with a set of tracked objects;
combining the tracking data associated with the set of tracked objects with the enriched data frame data of the current scene to create a combined dataset;
providing the combined dataset for use by the sensing apparatus;
identifying a first variance to assign to the tracking data; and
identifying a second variance to assign the enriched data frame data.

2. The method of claim 1 , wherein the first variance corresponds to a first estimated error associated with the tracking data and the second variance corresponds to a second estimated error associated with the enriched frame data.

3. The method of claim 1 , further comprising:
assigning a first weighting factor to the tracking data based on the first variance; and
assigning a second weighting factor to the enriched data frame data based on the second variance.

4. The method of claim 3 , wherein the first weighting factor and the second weighting factor are assigned values that are respectively inversely proportional to the first variance and the second variance, and processing of data points of the tracking data in combination with points of the enriched data frame data correspond to the values assigned to the first weighting factor and the second weighting factor.

5. The method of claim 1 , further comprising:
identifying a number of the points of the tracking data that are associated with a tracked object;
identifying a number of points of the enriched data frame data that are associated with the tracked object;
assigning a first weighting factor to assign to the points of the tracking data associated with the tracked object; and
assigning a second weighting factor to assign to the points of the enriched data frame data associated with the tracked object.

6. The method of claim 5 , wherein the first weighting factor and the second weighting factor are assigned values that are respectively proportional to the number of points of the tracking data that are associated with the tracked object and to the number of points of the enriched data frame data associated with the tracked object.

7. The method of claim 1 , further comprising:
identifying that a pattern associated with a set of data points of the enriched data frame data is inconsistent with patterns associated with previously observed objects; and
assigning a quality metric to the set of data points associated with the pattern based on the pattern being inconsistent with the patterns associated with the previously observed objects.

8. The method of claim 1 , further comprising identifying based on a priority of a quality metric associated with the enriched data frame data at least one of new data to include in the combined data set or redundant data that should not be included in the combined data set.

9. A non-transitory computer-readable storage medium having embodied thereon a program for implementing a method for improving perception of a sensing apparatus, the method comprising:
receiving radar sensor data associated with the current scene;
generating enriched data frame data of the current scene from the received radar sensor data;
accessing tracking data associated with a set of tracked objects;
combining the tracking data associated with the set of tracked objects with the enriched data frame data of the current scene to create a combined dataset; providing the combined dataset for use by the sensing apparatus; and
identify based on a priority of a quality metric associated with the enriched data frame data at least one of new data to include in the combined data set or redundant data that should not be included in the combined data set.

10. The non-transitory computer-readable storage medium of claim 9 , the program further executable to:
identify a first variance to assign to the tracking data; and
identify a second variance to assign the enriched data frame data.

11. The non-transitory computer-readable storage medium of claim 10 , wherein the first variance corresponds to a first estimated error associated with the tracking data and the second variance corresponds to a second estimated error associated with the enriched frame data.

12. The non-transitory computer-readable storage medium of claim 10 , further comprising:
assigning a first weighting factor to the tracking data based on the first variance; and
assigning a second weighting factor to the enriched data frame data based on the second variance.

13. The non-transitory computer-readable storage medium of claim 12 , wherein the first weighting factor and the second weighting factor are assigned values that are respectively inversely proportional to the first variance and the second variance, and processing of data points of the tracking data in combination with points of the enriched data frame data correspond to the values assigned to the first weighting factor and the second weighting factor.

14. The non-transitory computer-readable storage medium of claim 10 , the program further executable to:
identify a number of the points of the tracking data that are associated with a tracked object;
identify a number of points of the enriched data frame data that are associated with the tracked object;
assigning a first weighting factor to assign to the points of the tracking data associated with the tracked object; and
assign a second weighting factor to assign to the points of the enriched data frame data associated with the tracked object.

15. The non-transitory computer-readable storage medium of claim 14 , wherein the first weighting factor and the second weighting factor are assigned values that are respectively proportional to the number of points of the tracking data that are associated with the tracked object and to the number of points of the enriched data frame data associated with the tracked object.

16. The non-transitory computer-readable storage medium of claim 9 , the program further executable to:
identify that a pattern associated with a set of data points of the enriched data frame data is inconsistent with patterns associated with previously observed objects; and
assign a quality metric to the set of data points associated with the pattern based on the pattern being inconsistent with the patterns associated with the previously observed objects.

17. An apparatus, the apparatus comprising:
a memory; and
a processor that executes instructions out of the memory to:
access received radar sensor data associated with the current scene,
generate enriched data frame data of the current scene from the received radar sensor data,
accessing tracking data associated with a set of tracked objects,
combine the tracking data associated with the set of tracked objects with the enriched data frame data of the current scene to create a combined dataset, and provide the combined dataset for further processing,
identify that a pattern associated with a set of data points of the enriched data frame data is inconsistent with patterns associated with previously observed objects, and
assign a quality metric to the set of data points associated with the pattern based on the pattern being inconsistent with the patterns associated with the previously observed objects.

18. The apparatus of claim 17 , wherein the processor exe

Assignees

• Gm Cruise Holdings Llc

Inventors

• Wang Shaogang
• Karrman Anton Mario Bongio
• Pierce Brian
• Lin Kotung
• Jiang Yun

Classifications

• G01S13/931 Primary

  of land vehicles · CPC title

• G06V20/56

  exterior to a vehicle by using sensors mounted on the vehicle · CPC title

• G06V10/80

  Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level (multimodal speaker identification or verification G10L17/10) · CPC title

• G01S13/862

  Combination of radar systems with sonar systems · CPC title

• G01S13/865

  Combination of radar systems with lidar systems · CPC title