Systems and methods for adaptive clutter removal from radar scans

1. A method of adaptively identifying clutter points representing static objects from a sensor data scan, the method comprising:

obtaining a plurality of sensor data scans captured, by a sensor unit associated with a vehicle, at a plurality of consecutive time instants while the vehicle is traveling;

identifying a respective set of target points from each of the plurality of sensor data scans;

determining a characteristic indicative of a trajectory pattern relating to one or more respective sets of the target points, wherein the characteristic indicative of the trajectory pattern is determined by computing an average Doppler velocity between locations of the one or more respective sets of target points; and

in response to determining that the characteristic satisfies a pre-defined condition, associating an indicator with the respective sets of target points as relating to one or more static objects in an environment in which the vehicle is situated.

2. The method of claim 1 , wherein the characteristic indicative of the trajectory pattern is determined by:

determining whether a magnitude of the average Doppler velocity is below a threshold, or whether a direction of the average Doppler velocity is random compared with a prior Doppler velocity determined based sensor data scans captured prior to the one or more sensor data scans.

3. The method of claim 1 , wherein the characteristic indicative of the trajectory pattern is determined by:

computing, based on an inertial coordinate system, a track of the one or more respective sets of target points across the one or more sensor data scans, and determining whether a direction of the track is uniform in the inertial coordinate system.

4. The method of claim 3 , wherein the determining whether the direction of the track is uniform in the inertial coordinate system further comprises:

sampling a plurality of points at consecutive time instants from the computed track; and

determining whether a direction parameter along the plurality of sampled points has changed more than a threshold number of times during the consecutive time instants.

5. The method of claim 1 , wherein the characteristic indicative of the trajectory pattern is determined by:

computing, based on an inertial coordinate system, an average displacement vector between a first set of target points from a first sensor data scan captured at a first time instant and a second set of target points from a second sensor data scan captured at a second time instant later than the first time instant; and

determining whether a magnitude of the average displacement vector is greater than a threshold.

6. The method of claim 1 , wherein the characteristic indicative of the trajectory pattern is determined by:

computing a respective point density for the respective set of target points; and

determining whether the respective point density is greater than a threshold.

7. The method of claim 1 , further comprising:

determining a first characteristic and a second characteristic relating to the one or more respective sets of target points from one or more sensor data scans taken at consecutive time instants; and

in response to determining that the first characteristic and the second characteristic have conflicting indications of whether the respective sets of target points relate to one or more static objects:

applying a pre-defined priority rule ranking the first characteristic and the second characteristic; and

adopting an indication associated with a respective characteristic that has a higher priority from the pre-defined priority rule.

8. The method of claim 1 , further comprising:

monitoring the characteristic indicative of the trajectory pattern with the one or more sensor data scans taken in real time while the vehicle is traveling along a route; and

in response to determining that the characteristic no longer satisfies the pre-defined condition, removing the indicator with the respective sets of target points as relating to the one or more static objects.

9. The method of claim 8 , further comprising:

determining, over a period of time, a percentage of time when the respective set of target points is associated with the indicator as relating to the one or more static objects; and

in response to determining that the percentage of time is less than a threshold, continuing monitoring the respective set of target points in subsequent sensor data scans.

10. The method of claim 9 , further comprising:

in response to determining that the percentage of time is greater than a threshold, removing target points corresponding to the respective set of target points in a subsequent sensor data scan.

11. A system comprising:

one or more non-transitory computer-readable storage media embodying instructions; and

one or more processors coupled to the storage media and operable to execute the instructions to:

obtain a plurality of sensor data scans captured, by a sensor unit associated with a vehicle, at a plurality of consecutive time instants while the vehicle is traveling;

identify a respective set of target points from each of the plurality of sensor data scans;

determine a characteristic indicative of a trajectory pattern relating to one or more respective sets of target points, wherein the characteristic indicative of the trajectory pattern is determined by computing an average Doppler velocity between locations of the one or more respective sets of target points; and

in response to determining that the characteristic satisfies a pre-defined condition, associate an indicator with the respective sets of target points as relating to one or more static objects in an environment at which the vehicle is situated.

12. The system of claim 11 , wherein the characteristic indicative of the trajectory pattern is determined by:

determining whether a magnitude of the average Doppler velocity is below a threshold, or whether a direction of the average Doppler velocity is random compared with a prior Doppler velocity determined based sensor data scans captured prior to the one or more sensor data scans.

13. The system of claim 11 , wherein the characteristic indicative of the trajectory pattern is determined by:

computing, based on an inertial coordinate system, a track of the one or more respective sets of target points across the one or more sensor data scans; and

determining whether a direction of the track is uniform in the inertial coordinate system.

14. The system of claim 11 , wherein the characteristic indicative of the trajectory pattern is determined by:

computing, based on an inertial coordinate system, an average displacement vector between a first set of target points from a first sensor data scan captured at a first time instant and a second set of target points from a second sensor data scan captured at a second time instant later than the first time instant; and

determining whether a magnitude of the average displacement vector is greater than a threshold.

15. The system of claim 11 , wherein the characteristic indicative of the trajectory pattern is determined by:

computing a respective point density for the respective set of target points; and

determining whether the respective point density is greater than a threshold.

16. The system of claim 11 , wherein the one or more processors are operable to execute the instructions further to:

determine a first characteristic and a second characteristic relating to the one or more respective sets of target points from one or more sensor data scans taken at consecutive time instants; and

in response to determining that the first characteristic and the second characteristic have conflicting indications of whether the respective sets of target points relate to one or more static objects:

apply a pre-defined priority rule ranking the first characteristic and the second characteristic; and

adopt an indication associated with a respective characteristic that has a higher priority from the pre-defined priority rule.

17. The system of claim 11 , wherein the one or more processors are operable to execute the instructions further to:

monitor the characteristic with the sensor data scans taken in real time while the vehicle is traveling along a route; and

in response to determining that the characteristic no longer satisfies the pre-defined condition, remove the indicator with the respective sets of target points as relating to the one or more static objects.

18. The system of claim 17 , wherein the one or more processors are operable to execute the instructions further to:

determine, over a period of time, a percentage of time when the respective set of target points is associated with the indicator as relating to the one or more static objects; and

in response to determining that the percentage of time is less than a threshold, continue monitoring the respective set of target points in subsequent sensor data scans.

19. The system of claim 18 , wherein the one or more processors are operable to execute the instructions further to:

in response to determining that the percentage of time is greater than a threshold, remove target points corresponding to the respective set of target points in a subsequent sensor data scan.

20. One or more non-transitory computer-readable storage media embodying software that is operable when executed by a computing system to:

obtain a plurality of sensor data scans captured, by a sensor unit placed on a vehicle, at a plurality of consecutive time instants while the vehicle is traveling along a route;

identify a respective set of target points from each of the plurality of sensor data scans;

determine a characteristic indicative of a trajectory pattern relating to one or more respective sets of target points, wherein the characteristic indicative of the trajectory pattern is determined by computing an average Doppler velocity between locations of the one or more respective sets of target points; and

in response to determining that the characteristic satisfies a pre-defined condition, associate an indicator with the respective sets of target points as relating to one or more static objects in an environment at which the vehicle is situated.