Vehicle path planning
A computing system can receive, in a vehicle, moving object information is determined by processing lidar sensor data acquired by a stationary lidar sensor. The moving object information can be determined using typicality and eccentricity data analysis (TEDA) on the lidar sensor data. The vehicle can be operated based on the moving object information.
1. A method, comprising:
receiving, in a vehicle, moving object data determined by processing lidar sensor data acquired by a stationary lidar sensor performing a scan of a field of view, and processed using typicality and eccentricity data analysis (TEDA), wherein the stationary lidar sensor acquires lidar sensor data in a sequence of columns from left to right and transmits the lidar sensor data to a traffic infrastructure computing device which processes the columns of lidar sensor data in an order of the sequence in which they are acquired to determine the moving object data, wherein a portion of the lidar sensor data including the moving object data is received in the vehicle before the stationary lidar sensor has completed the scan of the field of view; and
operating the vehicle based on the moving object data.
2. The method of claim 1 , wherein TEDA includes processing the stationary lidar sensor data to determine a pixel mean and a pixel variance over a moving time window and combining current pixel values with pixel mean and pixel variance to determine foreground pixels based on eccentricity.
3. The method of claim 2 , wherein determining moving object information is based on determining connected regions of foreground pixels in a foreground/background image formed by TEDA.
4. The method of claim 3 , wherein determining moving object information in the foreground/background image includes tracking connected regions of foreground pixels in a plurality of foreground/background images.
5. The method of claim 4 , wherein moving object information is projected onto a map centered on the vehicle based on a 3D lidar sensor pose and lidar sensor field of view and a 3D vehicle pose.
6. The method of claim 5 , wherein operating the vehicle includes determining a polynomial function that includes predicted vehicle trajectories, wherein predicted vehicle trajectories include location, direction, speed, and lateral and longitudinal accelerations.
7. The method of claim 6 , wherein determining the polynomial function includes determining a destination location on the map.
8. The method of claim 7 , wherein determining the polynomial function includes avoiding collisions or near-collisions with moving objects.
9. A system, comprising a processor; and a memory, the memory including instructions to be executed by the processor to:
receive, in a vehicle, moving object data determined by processing lidar sensor data acquired by a stationary lidar sensor performing a scan of a field of view, and processed using typicality and eccentricity data analysis (TEDA), wherein the stationary lidar sensor acquires lidar sensor data in a sequence of columns from left to right and transmits the lidar sensor data to a traffic infrastructure computing device which processes the columns of lidar sensor data in an order of the sequence in which they are acquired to determine the moving object data, wherein a portion of the lidar sensor data including the moving object data is received in the vehicle before the stationary lidar sensor has completed the scan of the field of view; and
operate the vehicle based on the moving object information.
10. The system of claim 9 , wherein TEDA includes processing the stationary lidar sensor data to determine a pixel mean and a pixel variance over a moving time window and combining current pixel values with pixel mean and pixel variance to determine eccentricity.
11. The system of claim 9 , wherein determining moving object information is based on determining connected regions of foreground pixels in a foreground/background image formed by TEDA.
12. The system of claim 11 , wherein determining moving object information in the foreground/background image includes tracking connected regions of foreground pixels in a plurality of foreground/background images.
13. The system of claim 12 , wherein moving object information is projected onto a map centered on the vehicle based on a 3D lidar sensor pose and lidar sensor field of view and a 3D vehicle pose.
14. The system of claim 13 , wherein operating the vehicle includes determining a polynomial function on the map that includes predicted vehicle trajectories, wherein predicted vehicle trajectories include location, direction, speed, and lateral and longitudinal accelerations.
15. The system of claim 14 , wherein determining the polynomial function includes determining a destination location on the map.
16. The system of claim 15 , wherein determining the polynomial function includes avoiding collisions or near-collisions with moving objects.
17. A system, comprising:
means for controlling vehicle steering, braking and powertrain; and
computer means for:
receiving, in a vehicle, moving object data determined by processing lidar sensor data acquired by a stationary lidar sensor performing a scan of a field of view, and processed using typicality and eccentricity data analysis (TEDA), wherein the stationary lidar sensor acquires lidar sensor data in a sequence of columns from left to right and transmits the lidar sensor data to a traffic infrastructure computing device which processes the columns of lidar sensor data in an order of the sequence in which they are acquired to determine the moving object data, wherein a portions of the lidar sensor data including the moving object data is received in the vehicle before the stationary lidar sensor has completed the scan of the field of view; and
means for operating the vehicle based on the moving object information and the means for controlling steering, braking and powertrain.
18. The system of claim 9 , wherein the sequential columns of lidar sensor data are included in a field of view and portions of the lidar sensor data including the moving object information are received in the vehicle before the stationary lidar sensor has completed acquiring the field of view.
19. The system of claim 9 , wherein an empirically determined constant learning rate is used to assign an exponentially decreasing weights to the pixels of the lidar sensor data.
20. The system of claim 9 , wherein operating the vehicle is based on a cognitive map of the environment determined based on moving object information.