Systems and methods for object detection with LiDAR decorrelation

1. A method for operating an autonomous vehicle, comprising:

obtaining, by a computing device, LiDAR data generated by a LiDAR system of the autonomous vehicle and image data generated by a camera of the autonomous vehicle;

performing, by the computing device, a first object detection algorithm to generate first object detection information using the LiDAR data, and a second object detection algorithm to generate second object detection information using both the LiDAR data and the image data;

processing, by the computing device, the first and second object detection information to determine whether a given object was detected by both the first and second object detection algorithms;

performing, by the computing device, the following operations when a determination is made that the given object was detected by both the first and second object detection algorithms;

selectively modifying the first object detection information based on contents of the second object detection information; and

using, by the computing device, the modified first object detection information to facilitate at least one autonomous driving operation.

2. The method according to claim 1 , further comprising using, by the computing device, the second object detection information to facilitate at least one autonomous driving operation, when a determination is made that the given object was not detected by both the first and second object detection algorithms.

3. The method according to claim 1 , wherein the first object detection algorithm comprises a closed world object detection algorithm implemented by a neural network that is trained to detect particular types of objects.

4. The method according to claim 3 , wherein the second object detection algorithm comprises an open world object detection algorithm that is configured to detect any type of object.

5. The method according to claim 4 , wherein the open world detection algorithm comprises camera-LiDAR fusion object detection.

6. The method according to claim 4 , wherein the first object detection information is selectively modified based on (a) a total number of data points N open of the LiDAR data that are contained in a point cloud segment S open defined for an object detected by the open world detection algorithm and (b) a total number of data points N closed of the LiDAR data that are contained in a point cloud segment S closed defined for an object detected by the closed world detection algorithm.

7. The method according to claim 6 , wherein a size of the point cloud segment S closed is increased when the total number of data points N open is greater than the total number of data points N closed .

8. The method according to claim 6 , wherein a size of the point cloud segment S closed is decreased when the total number of data points N open is less than the total number of data points N closed .

9. The method according to claim 1 , further comprising discarding the first object detection information such that the first object detection information is not used to facilitate the at least one autonomous driving operation.

10. The method according to claim 1 , further comprising:

determining whether the first object detection information should be modified based on a difference in content between the first and second object detection information;

wherein the first object detection information is selectively modified only when a determination is made that the first object detection information should be modified.

11. A system, comprising:

a processor;

a non-transitory computer-readable storage medium comprising programming instructions that are configured to cause the processor to implement a method for operating an autonomous vehicle, wherein the programming instructions comprise instructions to:

obtain LiDAR data generated by a LiDAR system of the autonomous vehicle and image data generated by a camera of the autonomous vehicle;

perform a first object detection algorithm to generate first object detection information using the LiDAR data, and a second object detection algorithm to generate second object detection information using both the LiDAR data and the image data;

process the first and second object detection information to determine whether a given object was detected by both the first and second object detection algorithms;

perform the following operations when a determination is made that the given object was detected by both the first and second object detection algorithms;

selectively modify the first object detection information based on contents of the second object detection information; and

causing the modified first object detection information to be used for at least one autonomous driving operation.

12. The system according to claim 11 , wherein the programming instructions further comprise instructions to cause the second object detection information to be used at least one autonomous driving operation, when a determination is made that the given object was not detected by both the first and second object detection algorithms.

13. The system according to claim 11 , wherein the first object detection algorithm comprises a closed world object detection algorithm implemented by a neural network that is trained to detect particular types of objects.

14. The system according to claim 13 , wherein the second object detection algorithm comprises an open world object detection algorithm that is configured to detect any type of object.

15. The system according to claim 14 , wherein the open world detection algorithm comprises camera-LiDAR fusion object detection.

16. The system according to claim 14 , wherein the first object detection information is selectively modified based on (a) a total number of data points N open of the LiDAR data that are contained in a point cloud segment S open defined for an object detected by the open world detection algorithm and (b) a total number of data points N closed of the LiDAR data that are contained in a point cloud segment S closed defined for an object detected by the closed world detection algorithm.

17. The system according to claim 16 , wherein a size of the point cloud segment S closed is increased when the total number of data points N open is greater than the total number of data points N closed .

18. The system according to claim 16 , wherein a size of the point cloud segment S closed is decreased when the total number of data points N open is less than the total number of data points N closed .

19. The system according to claim 11 , wherein the programming instructions further comprise instructions to discard the first object detection information such that the first object detection information is not used to facilitate the at least one autonomous driving operation.

20. The system according to claim 11 , wherein the programming instructions further comprise instructions to:

determine whether the first object detection information should be modified based on a difference in content between the first and second object detection information;

wherein the first object detection information is selectively modified only when a determination is made that the first object detection information should be modified.