Apparatus and method for clustering point cloud

1. An apparatus for clustering a point cloud, the apparatus comprising:

a three-dimensional (3D) light detection and ranging (LiDAR) sensor configured to generate a point cloud around a vehicle; and

a controller configured to project the point cloud generated by the 3D LiDAR sensor onto a circular grid map to be converted into two-dimensional (2D) points, the circular grid map including a plurality of cells, and to cluster the 2D points on the circular grid map,

wherein the controller is configured to:

detect a plurality of representative points for each cell of the circular grid map; and

determine whether to perform clustering between the respective representative points based on the representative point for each cell.

2. The apparatus of claim 1 , wherein:

the circular grid map includes a plurality of circles,

each of the plurality of circles has a different size,

each of the plurality of circles uses as an origin thereof a center of the vehicle,

each of the plurality of circles includes a plurality of straight lines connected from the origin to a circular arc, and

the plurality of straight lines are spaced apart from each other at a threshold angle.

3. The apparatus of claim 2 , wherein:

a first straight line configures a first cell generated by a first circle of the plurality of circles, the first circle being the smallest circle of the plurality of circles,

a second straight line configures a second cell generated by a second circle of the plurality of circles at an outer portion of the first circle,

a third straight line configures a third cell generated by a third circle of the plurality of circles at an outer portion of the second circle,

the first straight line is shorter in length than the second straight line, and

the second straight line is shorter in length than the third straight line.

4. The apparatus of claim 1 , wherein the controller is configured to:

set a plurality of reference points for each cell of the circular grid map; and

detect a 2D point located closest to each of the plurality of reference points as the representative point for each cell.

5. The apparatus of claim 4 , wherein the plurality of reference points are located on a boundary line of a given cell of the plurality of cells.

6. The apparatus of claim 1 , wherein the controller is configured to:

cluster a representative point of a first cell of the plurality of cells and a representative point of a second cell of the plurality of cells when the representative point of the first cell and the representative point of the second cell are included in a reference frame.

7. The apparatus of claim 6 , wherein:

a shape and a size of the reference frame correspond to a shape and a size of the first cell, or

the shape and the size of the reference frame correspond to a shape and size of the second cell.

8. The apparatus of claim 1 , wherein the controller is configured to:

recognize one cluster as one object.

9. The apparatus of claim 8 , wherein the object is a vehicle.

10. A method for clustering a point cloud, the method comprising:

generating, by a 3D LiDAR sensor of a vehicle, a point cloud around a vehicle;

projecting, by a controller of the vehicle, the point cloud onto a circular grid map to be converted into two-dimensional (2D) points, the circular grid map including a plurality of cells; and

clustering, by the controller, the 2D points on the circular grid map,

wherein the clustering of the 2D points on the circular grid map comprises:

detecting a plurality of representative points for each cell of the circular grid map; and

determining whether to perform clustering between the respective representative points based on the representative point for each cell.

11. The method of claim 10 , wherein:

the circular grid map includes a plurality of circles,

each of the plurality of circles has a different size,

each of the plurality of circles uses as an origin thereof a center of the vehicle,

each of the plurality of circles includes a plurality of straight lines connected from the origin to a circular arc, and

the plurality of straight lines are spaced apart from each other at a threshold angle.

12. The method of claim 11 , wherein:

a first straight line configures a first cell generated by a first circle of the plurality of circles, the first circle being the smallest circle of the plurality of circles,

a second straight line configures a second cell generated by a second circle of the plurality of circles at an outer portion of the first circle,

a third straight line configures a third cell generated by a third circle of the plurality of circles at an outer portion of the second circle,

the first straight line is shorter in length than the second straight line, and

the second straight line is shorter in length than the third straight line.

13. The method of claim 10 , wherein the detecting of the plurality of representative points comprises:

setting a plurality of reference points for each cell of the circular grid map; and

detecting a 2D point located closest to each of the plurality of reference points as the representative point for each cell.

14. The method of claim 13 , wherein the plurality of reference points are located on a boundary line of a given cell of the plurality of cells.

15. The method of claim 10 , wherein the determining of whether to perform the clustering comprises:

clustering a representative point of a first cell of the plurality of cells and a representative point of a second cell of the plurality of cells when the representative point of the first cell and the representative point of the second cell are included in a reference frame.

16. The method of claim 15 , wherein:

a shape and a size of the reference frame correspond to a shape and a size of the first cell, or

the shape and the size of the reference frame correspond to a shape and size of the second cell.

17. The method of claim 10 , further comprising:

recognizing, by the controller, one cluster as one object.

18. The method of claim 17 , wherein the object is a vehicle.