Adaptive search for LiDAR-based clustering

1. A method of clustering spatial data, the method comprising:

receiving a point cloud comprised of a plurality of points defined within three-dimensional (3D) space;

selecting one or more adaptable clustering parameters for each of the plurality of points; and

traversing each of the plurality of points in the point cloud and selectively adding each of the points to one or more clusters based on the selected clustering parameters associated with each point.

2. The method of claim 1 , wherein selecting one or more adaptable clustering parameters includes selecting a minimum point threshold parameter or a search radius parameter.

3. The method of claim 2 , wherein the minimum point threshold parameter or search radius parameter is selected based on a distance to the point.

4. The method of claim 3 , wherein the minimum point threshold parameter is decreased as the determined distance to the point increases.

5. The method of claim 3 , wherein the search radius is increased as the determined distance to the point increases.

6. The method of claim 1 , further including:

calculating a density associated with each of the plurality of points based on a number of neighboring points located within a search radius of the point.

7. The method of claim 6 , wherein calculating the density associated with each of the plurality of points includes:

identifying the point as a core point based on a comparison of a number of neighboring points located within a search radius of the point with the defined minimum point threshold; and

increasing the search radius if the point is not identified as a core point and retrieving additional neighboring points based on the increased search radius, wherein calculating the density of the point is based on the increased search radius and additional neighboring points.

8. The method of claim 6 , further including

sorting the plurality of points based on the calculated density of each point.

9. The method of claim 8 , wherein traversing each of the plurality of points in the point cloud includes traversing points based on the sorted density of the plurality of points, wherein points are traversed from dense to sparse.

10. The method of claim 1 , wherein traversing each of the plurality of points in the point cloud and selectively adding each of the points to one or more clusters includes:

selecting a previously unclustered point;

creating a new cluster if the selected point is a core point and previously unclustered;

adding neighboring points to a neighbor list associated with the new cluster if the point is a core point, wherein neighboring points are located within a search radius of the core point; and

reviewing each point on the neighbor list and identifying each point as a core point or a border point, wherein a neighbor point identified as a core point is assigned to the new cluster and neighbors of the neighbor point identified as a core point are added to the neighbor list, wherein a neighbor point identified as a border point is assigned to the new cluster but neighbors of the border point are not added to the neighbor list.

11. The method of claim 10 , wherein in response to identification of a point as a border point, incident angles between the border point and points adjacent to the border point are compared to a threshold, wherein if the difference is less than the threshold the adjacent point is added to the neighbor list associated with the new cluster if not previously assigned to a cluster.

12. The method of claim 11 , wherein the incident angles are azimuth angles or elevation angles.

13. A method for clustering spatial data, the method comprising:

receiving a point cloud comprised of a plurality of points defined within three-dimensional (3D) space;

selecting a previously unclustered point from the point cloud;

creating a new cluster if the selected point was previously unclustered and is determined to be a core point based on a number of neighboring points located within a search radius;

adding neighboring points of the selected point to a neighbor list;

for each point on the neighbor list, determining if the neighbor point is a core point;

for a neighboring point identified as a core point, adding the neighboring point to the cluster; and

for neighboring points not identified as a core point, determining if an azimuth or elevation angle between the neighboring point and an adjacent point are less than a threshold, wherein if the azimuth or elevation angle between the neighboring point and the adjacent point is less than a threshold the adjacent point is added to the neighbor list if not previously assigned to a cluster.

14. The method of claim 13 , wherein if the azimuth or elevation angle between the neighboring point and the adjacent point is greater than a threshold then the neighboring point is identified as a border point and added to the cluster and no further action is taken with respect to the adjacent point.

15. The method of claim 13 , further including determining whether the adjacent point has been visited, wherein the adjacent point is added to the neighbor list if the adjacent point has not been visited.

16. The method of claim 15 , wherein if the adjacent point has been visited and the adjacent point was previously determined to be a noise point, then the adjacent point is identified as a border point and added to the cluster.

17. The method of claim 15 , wherein if the adjacent point has been visited and the adjacent point was not previously determined to be a noise point, then the adjacent point is not added to the current cluster.

18. A system comprising:

at least one processor; and

a clustering module configured to, when executed by the at least one processor, implement a density-based clustering algorithm that clusters a plurality of points received by the system into one or more clusters, the density-based clustering algorithm configured to:

receive a point cloud comprised of a plurality of points defined within three-dimensional (3D) space;

apply a search algorithm to the point cloud, wherein the search algorithm selects clustering parameters including one or more of minimum point threshold and search radius for each point in the point cloud; and

apply a clustering algorithm to the point cloud, wherein the clustering algorithm traverses at least some of the points in the point cloud and utilizes the minimum point threshold and search radius associated with each point by the search algorithm to selectively add points to clusters.

19. The system of claim 18 , wherein for each point in the point cloud, applying the search algorithm includes:

determining a distance to a point; and

selecting a minimum point threshold for the point based on the determined distance to the point.

20. The system of claim 18 , wherein applying the search algorithm further includes:

identifying the point as a core point based on a comparison of a number of neighboring points located within the defined search radius of the point with the defined minimum point threshold; and

increasing the search radius if the point is not identified as a core point and retrieving additional neighboring points based on the increased search radius.

21. The system of claim 18 , wherein applying the search algorithm further includes calculating a density associated with each of the points based on a number of neighbors located within the selected search radius associated with each point.

22. The system of claim 21 , wherein applying the clustering algorithm further includes traversing at least some of the points in the point cloud based on calculated density associated with each point, wherein the points are traversed from dense to sparse.

23. The system of claim 18 , wherein for each point in the point cloud, applying the clustering algorithm further includes:

selecting a previously unclustered point;

creating a new cluster if the selected point is a core point and previously unclustered;

adding neighboring points to a neighbor list associated with the new cluster if the point is a core point, wherein neighboring points are located within a search radius of the core point; and

reviewing each point on the neighbor list and identifying each point as a core point or a border point, wherein a neighbor point identified as a core point is assigned to the new cluster and neighbors of the neighbor point identified as a core point are added to the neighbor list, wherein a neighbor point identified as a border point is assigned to the new cluster but neighbors of the border point are not added to the neighbor list.