Systems and methods for radar localization in autonomous vehicles

1. A localization method comprising:

receiving sensor data relating to an environment of a vehicle, the sensor data including a plurality of sensor returns associated with objects in the environment, each of the sensor returns having a plurality of corresponding attributes;

constructing a first plurality of sensor data groups, each including a self-consistent subset of the plurality of sensor returns based on a machine learning clustering method and their corresponding attributes;

defining, for each of the first plurality of sensor data groups, a first set of features, wherein the first set of features includes a histogram associated with each of the corresponding attributes, and wherein each group has an associated feature location;

determining, with a processor, a feature correlation between the first set of features and a second, previously determined set of features based on the histograms and the feature location; and

estimating a position of the vehicle based on the feature correlation.

2. The method of claim 1 , wherein the plurality of corresponding attributes includes at least one of Doppler shift, return power, and neighborhood similarity.

3. The method of claim 1 , wherein the sensor data includes at least radar data.

4. The method of claim 1 , wherein the first set of features includes a convex hull of the histogram.

5. The method of claim 1 , wherein the first set of features includes a summary statistic of one of the corresponding attributes.

6. The method of claim 1 , wherein determining the feature correlation includes performing an outlier removal procedure with respect to the first and second sets of features.

7. The method of claim 6 , wherein the outlier removal procedure is a random sample consensus (RANSAC) procedure.

8. The method of claim 1 , further including classifying each of the sensor data groups as being associated with one of a dynamic object, a static-moveable object, and a static-nonmoveable object, and determining the feature correlation based only on the sensor data groups associated with static-nonmoveable objects.

9. A system for controlling a vehicle, comprising:

a feature determination module, including a processor, configured to: receive sensor data relating to an environment of a vehicle, the sensor data including a plurality of sensor returns associated with objects in the environment, each of the sensor returns having a plurality of corresponding attributes; construct a first plurality of sensor data groups, each including a self-consistent subset of the plurality of sensor returns based on a machine learning clustering method and their corresponding attributes; and define, for each of the first plurality of sensor data groups, a first set of features, the first set of features includes a histogram associated with each of the corresponding attributes, and wherein each group has an associated feature location; and

a feature correlation module configured to determine, with a processor, a feature correlation between the first set of features and a second, previously determined set of features based on the histograms and the feature location.

10. The system of claim 9 , wherein:

the plurality of corresponding attributes includes at least one of Doppler shift, return power, and neighborhood similarity; and

the sensor data is at least one of radar data and lidar data.

11. The system of claim 9 , wherein the first set of features includes a summary statistic of one of the corresponding attributes.

12. The system of claim 9 , wherein the feature correlation module performs an outlier removal procedure with respect to the first and second sets of features.

13. The system of claim 12 , wherein the outlier removal procedure is a random sample consensus (RANSAC) procedure.

14. The system of claim 9 , wherein the feature determination module classifies each of the sensor data groups as being associated with one of a dynamic object, a static-moveable object, and a static-nonmoveable object, and the feature correlation module determines the feature correlation based only on the sensor data groups associated with static-nonmoveable objects.

15. An autonomous vehicle, comprising:

at least one sensor that provides sensor data relating to an environment of the autonomous vehicle, the sensor data including a plurality of sensor returns associated with objects in the environment, each of the sensor returns having a plurality of corresponding attributes; and

a controller that, by a processor:

receives the sensor data;

constructs a first plurality of sensor data groups, each including a self-consistent subset of the plurality of sensor returns based on a machine learning clustering method and their corresponding attributes;

defines, for each of the first plurality of sensor data groups, a first set of features, wherein the first set of features includes a histogram associated with each of the corresponding attributes, and wherein each group has an associated feature location;

determines, with a processor, a feature correlation between the first set of features and a second, previously determined set of features based on the histograms and the feature location; and

estimates a position of the vehicle based on the feature correlation.

16. The autonomous vehicle of claim 15 , wherein the controller performs outlier removal via a random sample consensus (RANSAC) procedure to determine the feature correlation.

17. The autonomous vehicle of claim 15 , wherein

the plurality of corresponding attributes includes at least one of Doppler shift, return power, and neighborhood similarity; and

the sensor data includes radar data.