Sequential clustering

1. A method comprising, by a computing system associated with a vehicle:

accessing a plurality of data points captured using a radar system of the vehicle, wherein each data point is associated with at least three measurements comprising a Doppler measurement, a range measurement, and an azimuth measurement in reference to the radar system;

clustering the plurality of data points into one or more first clusters based on a first pair of the three measurements associated with each of the data points;

clustering the plurality of data points into one or more second clusters based on a second pair of the three measurements associated with each of the data points, the second pair being different from the first pair of the three measurements, wherein the one or more first clusters and the one or more second clusters each include a plurality of the data points;

comparing the one or more first clusters and the one or more second clusters to identify one or more overlap clusters that represent respective ones of the first clusters and the second clusters that correspond,

wherein comparing includes determining an intersection of data points between a particular first cluster of the first clusters and a particular second cluster of the second clusters using dimensions corresponding to the second pair, and wherein comparing includes determining a center of an overlap cluster of the overlap clusters by weighting data points of the overlap cluster according to a received signal strength indicator (RSSI) of the data points;

associating the data points of each of the overlap clusters as an object in an environment of the vehicle; and

removing one or more of the overlap clusters that have been previously identified based on a number of data points of the one or more of the overlap clusters being less than a threshold number of data points.

2. The method of claim 1 , wherein comparison comprises:

identifying the particular second cluster corresponding to the particular first cluster based on a similar range associated with the data points of the particular first cluster and the data points of the particular second cluster.

3. The method of claim 2 , further comprising assigning one or more data points of the particular first cluster or the particular second cluster outside the intersection of the plurality of data points to a corresponding overlap cluster based on a value of the RSSI of the one or more data points being higher than a threshold RSSI value.

4. The method of claim 1 , further comprising removing one or more data points from one or more of the overlap clusters based on a Doppler variation of the one or more data points being greater than a threshold amount of Doppler variation.

5. The method of claim 1 , further comprising removing one or more data points from one or more of the overlap clusters based on a range variation of the one or more data points being greater than a threshold amount of range variation.

6. The method of claim 1 , further comprising removing one or more data points from one or more of the overlap clusters based on an azimuth variation of the one or more data points being greater than a threshold amount of azimuth variation.

7. The method of claim 1 , further comprising filtering one or more data points in a range/azimuth space based on a value of the RSSI of the data point being lower than a threshold RSSI value.

8. The method of claim 1 , further comprising determining a center location associated with the one or more overlap clusters based on an average of a value of each location of the data points of the overlap cluster.

9. The method of claim 1 , further comprising determining a Doppler velocity associated with the one or more overlap clusters based on an average Doppler velocity of the data points of the overlap cluster and weighted by a value of the RSSI of each data point of the overlap cluster.

10. The method of claim 1 , further comprising filtering data points that are not associated with one of the one or more overlap clusters.

11. One or more computer-readable non-transitory storage media embodying software that is operable when executed to cause one or more processors to perform operations comprising:

accessing a plurality of data points captured using a radar system of a vehicle, wherein each data point is associated with at least three measurements comprising a Doppler measurement, a range measurement, and an azimuth measurement in reference to the radar system;

clustering the plurality of data points into one or more first clusters based on a first pair of the three measurements associated with each of the data points;

clustering the plurality of data points into one or more second clusters based on a second pair of the three measurements associated with each of the data points, the second pair being different from the first pair of the three measurements, wherein the one or more first clusters and the one or more second clusters each include a plurality of the data points;

comparing the one or more first clusters and the one or more second clusters to identify one or more overlap clusters that represent respective ones of the first clusters and the second clusters that correspond,

wherein comparing includes determining an intersection of data points between a particular first cluster of the first clusters and a particular second cluster of the second clusters using dimensions corresponding to the second pair, and wherein comparing includes determining a center of an overlap cluster of the overlap clusters by weighting data points of the overlap cluster according to a received signal strength indicator (RSSI) of the data points;

associating the data points of each of the overlap clusters as an object in an environment of the vehicle; and

removing one or more of the overlap clusters that have been previously identified based on a number of data points of the one or more of the overlap clusters being less than a threshold number of data points.

12. The computer-readable non-transitory storage media of claim 11 , wherein the software is further configured to:

identify the particular second cluster corresponding to the particular first cluster based on a similar range associated with the data points of the particular first cluster and the data points of the particular second cluster.

13. The computer-readable non-transitory storage media of claim 12 , wherein the software is further configured to assign one or more data points of the particular first cluster or the particular second cluster outside the intersection of the plurality of data points to a corresponding overlap cluster based on a value of the RSSI of the one or more data points being higher than a threshold RSSI value.

14. The computer-readable non-transitory storage media of claim 11 , wherein the software is further configured to remove one or more data points from one or more of the overlap clusters based on a Doppler variation of the one or more data points being greater than a threshold amount of Doppler variation.

15. A computing system comprising:

one or more processors; and

one or more computer-readable non-transitory storage media coupled to one or more of the processors, the one or more computer-readable non-transitory storage media comprising instructions operable when executed by one or more of the processors to cause the computing system to perform operations comprising:

accessing a plurality of data points captured using a radar system of a vehicle, wherein each data point is associated with at least three measurements comprising a Doppler measurement, a range measurement, and an azimuth measurement in reference to the radar system;

clustering the plurality of data points into one or more first clusters based on a first pair of the three measurements associated with each of the data points;

clustering the plurality of data points into one or more second clusters based on a second pair of the three measurements associated with each of the data points, the second pair being different from the first pair of the three measurements;

comparing the one or more first clusters and the one or more second clusters to identify one or more overlap clusters that represent respective ones of the first clusters and the second clusters that correspond, wherein the one or more first clusters and the one or more second clusters each include a plurality of the data points,

wherein comparing includes determining an intersection of data points between a particular first cluster of the first clusters and a particular second cluster of the second clusters using dimensions corresponding to the second pair, and wherein comparing includes determining a center of an overlap cluster of the overlap clusters by weighting data points of the overlap cluster according to a received signal strength indicator (RSSI) of the data points;

associating the data points of each of the overlap clusters as an object m an environment of the vehicle; and

removing one or more of the overlap clusters that have been previously identified based on a number of data points of the one or more of the overlap clusters being less than a threshold number of data points.

16. The computing system of claim 15 , wherein the processors are further operable to:

identify the particular second cluster corresponding to the particular first cluster based on a similar range associated with the data points of the particular first cluster and the data points of the particular second cluster.

17. The computing system of claim 16 , wherein the processors are further operable to assign one or more data points of the particular first cluster or the particular second cluster outside the intersection of the plurality of data points to a corresponding overlap cluster based on a value of the RSSI of the one or more data points being higher than a threshold RSSI value.

18. The computing system of claim 15 , wherein the processors are further operable to remove one or more data points from one or more of the overlap clusters based on a Doppler variation of the one or more data points being greater than a threshold amount of Doppler variation.