Using obstacle clearance to measure precise lateral gap
A system and method is provided for identifying an object along a road, where the object may be represented by a bounding box, and projecting a set of obstacle points within the bounding box corresponding to the identified object. In one aspect, a two-dimensional plane oriented perpendicular to a direction of the movement of the vehicle may be identified. In another aspect, the areas of the plane that may be occupied based on the set of obstacle points may be determined to generate a contour of the identified object. Thereafter, the height profiles of the identified object and the vehicle may be determined and identified, respectively. Based on the height profiles, a minimum clearance may be determined.
1. A method maneuvering a vehicle in an autonomous driving mode alongside of an object, the method comprising:
receiving, by one or more computing devices having one or more processors, sensor data including three-dimensional points on one or more surfaces of the object in an environment of the vehicle;
determining, by the one or more computing devices, a two-dimensional contour of the object using the three-dimensional points;
determining, by the one or more computing devices, a plurality of horizontal clearance distances between the vehicle and the two-dimensional contour of the object; and
identifying, by the one or more computing devices, a smallest horizontal clearance distance of the plurality of horizontal clearance distances;
maneuvering, by the one or more computing devices, the vehicle both under at least a portion of the object and alongside at least a portion of the object based on at least a comparison of the smallest horizontal clearance distance and a threshold minimum horizontal distance.
2. The method of claim 1 , wherein the two-dimensional contour of the object corresponds to a curvature of the object in a plane perpendicular to a direction of movement of the vehicle.
3. The method of claim 1 , wherein determining the two-dimensional contour includes projecting the three-dimensional points onto a grid.
4. The method of claim 3 , wherein the grid corresponds to a plane perpendicular to a direction of movement of the vehicle.
5. The method of claim 3 , wherein determining the two-dimensional contour includes identifying occupied cells of the grid, where an occupied cell corresponds to a cell of the grid in which at least one of the three-dimensional points is projected.
6. The method of claim 5 , wherein determining the two-dimensional contour includes using the identified occupied cells to form the two-dimensional contour.
7. The method of claim 1 , wherein determining the plurality of horizontal clearance distances is based at least in part on a virtual model of the vehicle defining a cross-sectional area of the vehicle.
8. The method of claim 7 , wherein determining the plurality of horizontal clearance distances includes determining a height profile of the object as a set of horizontal distances between a center line of the virtual model of the vehicle and the two-dimensional contour.
9. The method of claim 7 , wherein determining the plurality of horizontal clearance distances includes determining a set of difference values between (1) a set of horizontal distances from a center line of the virtual model of the vehicle to an edge of the virtual model of the vehicle and (2) a set of horizontal distances between a center line of the virtual model of the vehicle and the two-dimensional contour.
10. The method of claim 7 , wherein maneuvering the vehicle alongside of the object is further based on the smallest horizontal clearance distance.
11. The method of claim 10 , further comprising determining that the smallest horizontal clearance distance meets the threshold minimum horizontal distance before using the smallest horizontal clearance distance to maneuver the vehicle.
12. The method of claim 11 , wherein the sensor data includes a three-dimensional bounding box which includes the three-dimensional points and the method further comprises, when the minimum horizontal clearance distance does not meet the threshold minimum horizontal distance, maneuvering the vehicle alongside of the object includes maneuvering the vehicle around the bounding box.
13. The method of claim 1 , wherein the sensor data includes a three-dimensional bounding box which includes the three-dimensional points, and wherein maneuvering the vehicle includes maneuvering the vehicle alongside of the object through a location corresponding to a point within a three-dimensional volume of the bounding box.
14. A system for maneuvering a vehicle in an autonomous driving mode alongside of an object, the system comprising one or more computing devices having one or more processors, the one or more computing devices being configured to:
receive sensor data including three-dimensional points on one or more surfaces of the object in an environment of the vehicle;
determine a two-dimensional contour of the object using the three-dimensional points; and
determine a plurality of horizontal clearance distances between the vehicle and the two-dimensional contour of the object; and
identify a smallest horizontal clearance distance of the plurality of horizontal clearance distances;
maneuver the vehicle both under at least a portion of the object and alongside at least a portion of the object based on at least a comparison of the smallest horizontal clearance distance and a threshold minimum horizontal distance.
15. The system of claim 14 , wherein the two-dimensional contour of the object corresponds to a curvature of the object in a plane perpendicular to a direction of movement of the vehicle.
16. The system of claim 14 , wherein the one or more processors are further configured to determine the two-dimensional contour by projecting the three-dimensional points onto a grid.
17. The system of claim 16 , wherein the grid corresponds to a plane perpendicular to a direction of movement of the vehicle.
18. The system of claim 16 , wherein the one or more processors are further configured to determine the two-dimensional contour by identifying occupied cells of the grid, where an occupied cell corresponds to a cell of the grid in which at least one of the three-dimensional points is projected.
19. The system of claim 18 , wherein the one or more processors are further configured to determine the two-dimensional contour includes by using the identified occupied cells to form the two-dimensional contour.
20. The system of claim 14 , further comprising the vehicle.