Systems and methods for using road understanding to constrain radar tracks

Patent metadata

Field	Value
Publication number	US-10935652-B2
Application number	US-201816018804-A
Country	US
Kind code	B2
Filing date	Jun 26, 2018
Priority date	Jun 26, 2018
Publication date	Mar 2, 2021
Grant date	Mar 2, 2021

Abstract

A processor-implemented method in a vehicle for detecting objects from radar data includes: retrieving radar measurements taken at different periodic time increments; organizing the radar measurements into appropriate time windows; building a sequence cluster of radar measurements wherein the sequence cluster comprises a sliding window the latest time windows of radar measurements; removing noise from the sequence cluster of radar measurements by removing a cluster of radar measurements from the sequence cluster of radar measurements that is contradictory to a road topology map for an area in which the first object is estimated to be situated; and outputting the sequence cluster of radar measurements after removal of contradictory radar measurements as a new cluster of radar measurements.

First claim

What is claimed is:

1. A processor-implemented method in a vehicle for detecting and tracking objects using radar data, the method comprising:
tracking one or more radar tracks using a separate instance of a constrained filter for each radar track, each radar track comprising consecutive observations of the same object, the tracking comprising enforcing and constraining the movement of the one or more radar tracks within an aligned lane in a manner the aligned lane dictates even when a measurement supporting a movement is missing from the one or more radar tracks;
determining if a new cluster of radar measurements corresponds to a detected object aligned within a first lane by comparing the new cluster of radar measurements to the radar tracks that are aligned within the first lane;
when the comparison results in the identification of a radar track to which the new cluster corresponds, searching in aligned lanes, which are different from the first lane, for earlier radar measurements corresponding to the identified radar track that are consistent with the new cluster of radar measurements; and
associating the new cluster of radar measurements with the identified radar track when the earlier radar measurements corresponding to the identified radar track are consistent with the new cluster of radar measurements.

2. The method of claim 1 , further comprising predicting a future observation for the identified radar track by projecting the future observation along a path dictated by the aligned lane.

3. The method of claim 2 , wherein predicting a future observation occurs when an occluding object prevents the vehicle from receiving radar returns from sections of the aligned lane.

4. The method of claim 2 , wherein predicting a future observation comprises predicting that the detected object may cross another lane.

5. The method of claim 2 , wherein predicting a future observation comprises predicting that the detected object may turn into another lane due to requirements of the aligned lane.

6. The method of claim 5 , wherein the aligned lane comprises a turn only lane section.

7. The method of claim 5 , wherein the aligned lane comprises a merge only lane section.

8. The method of claim 5 , wherein the aligned lane comprises a no-turn lane section.

9. The method of claim 1 , further comprising generating the new cluster of radar measurements by removing a cluster of radar measurements that is contradictory to a road topology map.

10. The method of claim 9 , wherein removing a cluster of radar measurements that is contradictory to a road topology map comprises:
comparing the sequence cluster of radar measurements to the road topology map, the road topology map configured to identify lanes and allowed directions of travel in the identified lanes; and
removing radar measurements that indicate object movement in a direction that is contradictory to an allowed direction of travel in the lanes in the road topology map.

11. An object detection system in a vehicle for detecting and tracking objects using radar data, the system comprising a controller configured by programming instructions to:
track one or more radar tracks using a separate instance of a constrained filter for each radar track, each radar track comprising consecutive observations of the same object, the tracking comprising enforcing and constraining the movement of the one or more radar tracks within an aligned lane in a manner the aligned lane dictates even when a measurement supporting a movement is missing from the one or more radar tracks;
determine if a new cluster of radar measurements corresponds to a detected object aligned within a first lane by comparing the new cluster of radar measurements to the radar tracks that are aligned within the first lane;
when the comparison results in the identification of a radar track to which the new cluster corresponds, search in aligned lanes, which are different from the first lane, for earlier radar measurements corresponding to the identified radar track that are consistent with the new cluster of radar measurements; and
associate the new cluster of radar measurements with the identified radar track when the earlier radar measurements corresponding to the identified radar track are consistent with the new cluster of radar measurements.

12. The object detection system of claim 11 , further configured to predict a future observation for the identified radar track by projecting the future observation along a path dictated by the aligned lane.

13. The object detection system of claim 12 , further configured to predict a future observation when an occluding object prevents the vehicle from receiving radar returns from sections of the aligned lane.

14. The object detection system of claim 12 , further configured to predict that the detected object may cross another lane.

15. The object detection system of claim 12 , further configured to predict that the detected object may turn into another lane due to requirements of the aligned lane.

16. The object detection system of claim 15 , wherein the aligned lane comprises a turn only lane section.

17. The object detection system of claim 15 , wherein the aligned lane comprises a merge only lane section.

18. The object detection system of claim 15 , wherein the aligned lane comprises a no-turn lane section.

19. The object detection system of claim 11 , further configured to generate the new cluster of radar measurements by removing a cluster of radar measurements that is contradictory to a road topology map.

20. The object detection system of claim 19 , wherein to remove a cluster of radar measurements that is contradictory to a road topology map the object detection system is configured to:
compare the sequence cluster of radar measurements to the road topology map, wherein the road topology map is configured to identify lanes and allowed directions of travel in the identified lanes; and
remove radar measurements that indicate object movement in a direction that is contradictory to an allowed direction of travel in the lanes in the road topology map.

Assignees

• Gm Global Tech Operations Llc

Inventors

• Branson Elliot

Classifications

• G01S13/726 Primary

  Multiple target tracking · CPC title

• G01S2013/9322

  using additional data, e.g. driver condition, road state or weather data · CPC title

• G01S2013/93271

  in the front of the vehicles · CPC title

• G01S13/931 Primary

  of land vehicles · CPC title