Detection of overhanging objects

1. A method of detecting overhanging objects in an external environment of an autonomous vehicle, the method comprising:

identifying one or more floating obstacle candidates based on sensor data acquired from at least a forward portion of the external environment;

filtering out the identified one or more floating obstacle candidates based on one or more predefined parameters;

determining a driving maneuver for the autonomous vehicle at least partially based on a height clearance between the autonomous vehicle and floating obstacle candidates that remain after being filtered out; and

causing the autonomous vehicle to implement the determined driving maneuver.

2. The method of claim 1 , wherein the sensor data is a plurality of object data points, and wherein identifying one or more floating obstacle candidates based on sensor data acquired from at least a forward portion of the external environment includes:

sensing at least the forward portion of the external environment to acquire the plurality of object data points;

filtering out the acquired object data points to remove object data points located outside of one or more road boundaries of a current road of the autonomous vehicle; and

grouping the acquired object data points that remain after being filtered out into one or more obstacle candidates, wherein the one or more floating obstacle candidates are identified from the one or more obstacle candidates.

3. The method of claim 2 , wherein sensing at least a forward portion of the external environment to acquire the plurality of object data points is performed using one or more LIDAR sensors.

4. The method of claim 2 , wherein filtering out the acquired object data points to remove object data points located outside of one or more road boundaries of the current road of the autonomous vehicle includes:

locating the acquired object data points on a map, wherein the map includes the current road of the autonomous vehicle and the one or more road boundaries of the current road of the autonomous vehicle; and

filtering out any object data points located laterally outside of the one or more road boundaries or located laterally outside of a predetermined distance from the one or more road boundaries.

5. The method of claim 2 , wherein grouping the object data points that remain after being filtered out into one or more obstacle candidates includes point clustering the object data points that remain after being filtered out into one or more obstacle candidates.

6. The method of claim 1 , wherein filtering out the identified one or more floating obstacle candidates based on one or more predefined parameters includes:

sensing at least a forward portion of the external environment to acquire speed data of objects located therein;

associating the acquired speed data with the one or more floating obstacle candidates;

comparing the speed data associated with the one or more floating obstacle candidates to a predetermined speed threshold; and

filtering out any of the floating obstacle candidates with associated speed data that is above the predetermined speed threshold.

7. The method of claim 6 , wherein sensing at least a forward portion of the external environment to acquire speed data is performed using one or more RADAR sensors.

8. The method of claim 6 , further including:

determining whether any floating obstacle candidate with associated speed data that is below the predetermined speed threshold is a floating object or an occluded object; and

wherein, if the floating obstacle candidate is determined to be an occluded object, filtering out the occluded object so that it is not used in determining the driving maneuver for the autonomous vehicle at least partially based on the height clearance of the filtered out one or more floating obstacle candidates; and

wherein, if the floating obstacle candidate is determined to be a floating object, determining the driving maneuver for the autonomous vehicle at least partially based on the height clearance includes determining the driving maneuver for the autonomous vehicle at least partially based on the height clearance of the floating object.

9. The method of claim 1 , wherein determining the driving maneuver for the autonomous vehicle at least partially based on the height clearance between the autonomous vehicle and floating obstacle candidates that remain after being filtered out includes:

determining whether the autonomous vehicle will collide with one or more of the floating obstacle candidates that remain after being filtered out based at least on a current driving path of the autonomous vehicle and the height clearance between the autonomous vehicle and the floating obstacle candidates that remain after being filtered out,

wherein, if it is determined that the autonomous vehicle will collide with one or more of the floating obstacle candidates that remain after being filtered out, the determined driving maneuver includes a lateral movement of the autonomous vehicle,

wherein, if it is determined that the autonomous vehicle will not collide with one or more of the floating obstacle candidates that remain after being filtered out, the determined driving maneuver includes maintaining the current driving path of the autonomous vehicle.

10. The method of claim 9 , wherein the lateral movement of the autonomous vehicle is one of a lane shift or a lane change.

11. The method of claim 9 , wherein the sensor data is a plurality of object data points, further including:

projecting the object data points associated with floating obstacle candidates that remain after being filtered out onto a plane substantially perpendicular to a driving direction of the autonomous vehicle; and

forming a first enclosure to around the projected object data points, wherein the height clearance of the filtered out one or more floating obstacle candidates is based at least partially on a height clearance between the autonomous vehicle and the first enclosure.

12. The method of claim 11 , further including:

wherein, if it is determined that the autonomous vehicle will collide with one or more of the floating obstacle candidates that remain after being filtered out, filtering out the projected object data points that are located at a higher elevation than a height of the autonomous vehicle;

project the object data points that remain after being filtered out to a ground plane; and

forming a second enclosure around the projected object data points on the ground plane, wherein the lateral movement of the autonomous vehicle is based on the second enclosure.

13. A system for detecting overhanging objects in an external environment of an autonomous vehicle, the system comprising:

a sensor system configured to acquire sensor data of at least a forward portion of the external environment of the autonomous vehicle; and

a processor operatively connected to the sensor system, the processor being programmed to initiate executable operations comprising:

identifying one or more floating obstacle candidates based on the acquired sensor data of at least a forward portion of the external environment;

filtering out the identified one or more floating obstacle candidates based on one or more predefined parameters;

determining a driving maneuver for the autonomous vehicle at least partially based on a height clearance between the autonomous vehicle and of the floating obstacle candidates that remain after being filtered out; and

causing the autonomous vehicle to implement the determined driving maneuver.

14. The system of claim 13 , wherein the sensor data is a plurality of object data points, and wherein identifying one or more floating obstacle candidates based on sensor data acquired from at least a forward portion of the external environment includes:

sensing, using the sensor system, at least the forward portion of the external environment to acquire the plurality of object data points;

filtering out the acquired object data points to remove object data points located outside of one or more road boundaries of a current road of the autonomous vehicle; and

grouping the object data points that remain after being filtered out into one or more obstacle candidates, wherein the one or more floating obstacle candidates are identified from the one or more obstacle candidates.

15. The system of claim 14 , wherein filtering out the acquired object data points to remove object data points located outside of one or more road boundaries of a current road of the autonomous vehicle includes:

locating the object data points on a map, wherein the map includes the current road of the autonomous vehicle and the one or more road boundaries of the current road of the autonomous vehicle; and

filtering out object data points located laterally outside of the one or more road boundaries or located laterally outside of a predetermined distance from the one or more road boundaries.

16. The system of claim 14 , wherein the sensor system includes one or more LIDAR sensors, wherein the sensor data is a plurality of object data points, and wherein the one or more LIDAR sensors are configured to sense at least a forward portion of the external environment to acquire object data points.

17. The system of claim 14 , wherein the sensor system includes one or more RADAR sensors, wherein the one or more RADAR sensors are configured to sense at least a forward portion of the external environment to acquire speed data for one or more objects located therein.

18. The system of claim 17 , wherein filtering out the identified one or more floating obstacle candidates based on one or more predefined parameters includes:

associating the acquired speed data with the one or more floating obstacle candidates;

comparing the speed data associated with the one or more floating obstacle candidates to a predetermined speed threshold; and

filtering out any of the floating obstacle candidates with associated speed data that is above the predetermined speed threshold.

19. The system of claim 18 , further including:

determining whether any floating obstacle candidate with associated speed data that is below the predetermined speed threshold is a floating object or an occluded object; and

wherein, if the floating obstacle candidate is determined to be an occluded object, filtering out the occluded object so that it is not used in determining the driving maneuver for the autonomous vehicle at least partially based on the height clearance of the filtered out one or more floating obstacle candidates; and

wherein, if the floating obstacle candidate is determined to be a floating object, determining the driving maneuver for the autonomous vehicle at least partially based on the height clearance between the autonomous vehicle and any floating obstacle candidates that remain after being filtered out includes determining the driving maneuver for the autonomous vehicle at least partially based on the height clearance of the floating object.

20. A computer program product for detecting overhanging objects in an external environment of an autonomous vehicle, the computer program product comprising a non-transitory computer readable storage medium having program code embodied therein, the program code executable by a processor to perform a method comprising:

identifying one or more floating obstacle candidates based on sensor data acquired from at least a forward portion of the external environment;

filtering out the identified one or more floating obstacle candidates based on one or more predefined parameters;

determining a driving maneuver for the autonomous vehicle at least partially based on a height clearance between the autonomous vehicle and floating obstacle candidates that remain after being filtered out; and

causing the autonomous vehicle to implement the determined driving maneuver.