Methods and systems for inferring unpainted stop lines for autonomous vehicles

1. A method of inferring a stop line for a vehicle at an intersection, the method comprising:

by a processor of a vehicle, while the vehicle is moving along a reference path on a first road in a geographic area:

accessing map data for a map, the map data comprising data representing a plurality of roads in the geographic area, wherein the map data segments each of the roads into lane segments;

detecting from the accessed map data that a traffic control measure is located proximate an intersection that is ahead of the vehicle along the reference path;

performing first operations to detect a stop line for the detected traffic control measure that is painted on the first road before or at the intersection;

in response to no stop line for the detected traffic control measure being detected, performing the following second operations to obtain an inferred location of the stop line:

identifying, as a restricted lane, a lane segment of the first road that (i) includes an entry line proximate a location of the traffic control measure and (ii) extends into the intersection, wherein the entry line at least partially defines a boundary of the restricted lane;

identifying, as a nearest lane conflict, a lane segment of a second road intersecting the restricted lane;

determining whether the entry line of the restricted lane is within a target distance of the lane segment of the second road intersecting the restricted lane;

setting the inferred location of the stop line to a location of the entry line responsive to a determination being made that the entry line is within the target distance of the lane segment of the second road; and

performing the following third operations responsive to a determination being made that the entry line is not within the target distance from the lane segment of the second road:

initializing a location of an inferred stop line as the location of the entry line of the restricted lane;

advancing the location of the inferred stop line a distance away from the vehicle and towards the nearest lane conflict;

iteratively repeating the advancing until the location of the inferred stop line is at the target distance from a nearest boundary of the nearest lane conflict;

setting a final location of the inferred stop line as the location of the inferred stop line at completion of said iteratively repeating;

refining the final location of the inferred stop line based on at least one object perceived in an environment using sensor data, wherein the at least one object excludes any traffic control measure; and

setting the inferred location of the stop line as the refined final location of the inferred stop line.

2. The method of claim 1 , further comprising, by the processor of the vehicle, causing the vehicle to decelerate and stop at a location in the restricted lane at or before the inferred location of the stop line.

3. The method of claim 1 , wherein the third operations further comprise:

constructing a line segment as an intermediate stop line between a first set of points alongside boundaries of the restricted lane; and

iteratively advancing opposite endpoints of the line segment along the side boundaries until each endpoint is at the target distance from the nearest boundary of the nearest lane conflict, thus yielding the inferred location of the stop line.

4. The method of claim 3 , wherein iteratively advancing opposite endpoints of the line segment comprises:

advancing the opposite endpoints independently until each of the opposite endpoints reaches the target distance from the nearest boundary of the nearest lane conflict; and

forming the inferred stop line to intersect the restricted lane at the same angle as the nearest lane conflict.

5. The method of claim 1 , further comprising, by the processor of the vehicle:

detecting that a crosswalk crosses the first road before the nearest lane conflict of the intersection,

wherein the third operations further comprise constructing a line segment as an intermediate stop line across the restricted lane and iteratively advancing opposite endpoints of the line segment until each endpoint is a target distance from a nearest boundary of the crosswalk.

6. The method of claim 5 , wherein iteratively advancing the opposite endpoints of the line segment comprises:

advancing the opposite endpoints independently until each of the opposite endpoints reaches the target distance from the nearest boundary of the crosswalk; and

forming the inferred stop line to intersect the restricted lane at the same angle as the nearest boundary of the crosswalk.

7. The method of claim 1 , further comprising, by the processor of the vehicle:

in response to determining that no lane conflict is identified as appropriate to infer a stop line, setting the entry line as the stop line; and

causing the vehicle to stop at the entry line of the restricted lane.

8. The method of claim 1 , further comprising, by the processor of the vehicle:

determining whether a line-of-sight of a vision system of the vehicle to an intersecting road of the intersection is blocked; and

refining the target distance, in response to the line-of-sight being blocked,

wherein said third operations include:

advancing the location of the entry line until the refined target distance from the nearest boundary of the nearest lane conflict is reached to form the inferred stop line.

9. The method of claim 8 , further comprising, by the processor of the vehicle, repeating:

determining whether the line-of-sight of the vision system of the vehicle remains blocked;

refining the target distance, in response to the line-of-sight remaining blocked, and

advancing the entry line until the refined target distance from the nearest boundary of the nearest lane conflict is reached,

wherein the repeating continues until a first one of the line-of-sight is unblocked or the nearest boundary of the nearest lane conflict is reached.

10. The method of claim 1 , wherein identifying the nearest lane conflict comprises:

identifying a plurality of candidate lane segments into which the vehicle may enter when moving along the reference path to traverse the second road;

evaluating the plurality of candidate lane segments to determine which are appropriate nearest lane conflicts by:

removing, from the candidate lane segments, any lane segment that merges with an outlet of the restricted lane, and

removing, from the candidate lane segments, any lane segment that neighbors an inlet of the restricted lane;

selecting, from the appropriate nearest lane conflicts, the nearest lane segment into which the vehicle will enter when moving along the reference path to traverse the second road, as the nearest lane conflict.

11. A system for inferring a stop line for a vehicle at an intersection, the system comprising:

sensors configured to capture data about an environment that is proximate the vehicle;

a processor; and

a non-transitory computer readable medium comprising programming instructions that, when executed by the processor, will cause the processor to, while the vehicle is moving along a reference path on a first road in a geographic area:

access map data for a map, the map data comprising data representing a plurality of roads in the geographic area, wherein the map data segments each of the roads into lane segments;

detect from the accessed map data that a traffic control measure is positioned before an intersection that is ahead of the vehicle in the reference path;

perform first operations to detect a stop line for the detected traffic control measure that is painted on the first road before or at the intersection;

in response to no stop line being detected, perform the following second operations to obtain an inferred location of the stop line:

identifying, as a restricted lane, a lane segment of the first road that (i) includes an entry line proximate a location of the traffic control measure and (ii) extends into the intersection, wherein the entry line at least partially defines a boundary of the restricted lane;

identifying, as a nearest lane conflict, a lane segment of a second road intersecting the first road at the intersection;

determining whether the entry line of the restricted lane is within a target distance of the lane segment of the second road intersecting the restricted lane;

setting the inferred location of the stop line to a location of the entry line responsive to a determination being made that the entry line is within the target distance of the lane segment of the second road; and

performing the following third operations responsive to a determination being made that the entry line is not within the target distance from the lane segment of the second road:

initializing a location of an inferred stop line as the location of the entry line of the restricted lane;

advancing the location of the inferred stop line a distance away from the vehicle and towards the nearest lane conflict;

iteratively repeating the advancing until the location of the inferred stop line is at the target distance from a nearest boundary of the nearest lane conflict;

setting a final location of the inferred stop line as the location of the inferred stop line at completion of said iteratively repeating;

refining the final location of the inferred stop line based on at least one object perceived in an environment using sensor data, wherein the at least one object excludes any traffic control measure; and

setting the inferred location of the stop line as the refined final location of the inferred stop line.

12. The system of claim 11 , further comprising programming instructions that will cause the processor to cause the vehicle to decelerate and stop at a location in the restricted lane at or before the inferred location of the stop line.

13. The system of claim 11 , wherein the third operations further comprise:

constructing a line segment as an intermediate stop line between a first set of points alongside boundaries of the restricted lane; and

iteratively advance opposite endpoints of the line segment along the side boundaries until the endpoints are equidistant from the target distance from the nearest boundary of the nearest lane conflict, thus yielding the inferred location of the stop line.

14. The system of claim 13 , further comprising programming instructions that will cause the processor to, when iteratively advancing the opposite endpoints of the line segment:

advance the opposite endpoints independently until each of the opposite endpoints reaches the target distance from the nearest boundary of the nearest lane conflict, forming the inferred stop line to intersect the restricted lane at the same angle as the boundary of the nearest lane conflict.

15. The system of claim 11 , further comprising programming instructions that will cause the processor to:

detect that a crosswalk crosses the first road before the nearest lane conflict of the intersection,

construct a line segment as an intermediate stop line across the restricted lane; and

iteratively advance opposite endpoints of the line segment to a location that is equidistant from and at a target distance from a nearest boundary of the crosswalk.

16. The system of claim 15 , wherein the programming instructions that will cause the processor to iteratively advance the opposite endpoints of the line segment further comprise programming instructions that will cause the processor to:

advancing the opposite endpoints of the line segment independently until each of the opposite endpoints reaches the target distance from the nearest boundary of the crosswalk; and

forming the inferred stop line to intersect the restricted lane at the same angle as the nearest boundary of the crosswalk.

17. The system of claim 11 , further comprising programming instructions that will cause the processor to:

in response to determining that no nearest lane conflict is identified as appropriate to infer a stop line, set the entry line as the stop line; and

cause the vehicle to stop at the entry line of the restricted lane.

18. The system of claim 11 , further comprising programming instructions that will cause the processor to:

determine a line-of-sight of a vision system of the vehicle to an intersecting road of the intersection is blocked; and

refine the target distance, in response to the line-of-sight being blocked,

advance the location of the entry line until the refined target distance from the nearest boundary of the nearest lane conflict is reached to form the inferred stop line.

19. The system of claim 18 , further comprising programming instructions that will cause the processor to:

determine whether the line-of-sight of the vision system of the vehicle remains blocked;

refine the target distance, in response to the line-of-sight remaining blocked; and

advance the entry line until the refined target distance from the boundary of the nearest lane conflict is reached until the line-of-sight is unblocked.

20. The system of claim 11 , wherein the programming instructions that will cause the processor to identify the nearest lane conflict further comprise programming instructions that will cause the processor to:

identify a plurality of candidate lane segments into which the vehicle may enter when moving along the reference path to traverse the second road;

evaluate the plurality of candidate lane segments to determine which are appropriate nearest lane conflicts by:

removing, from the candidate lane segments, any lane segment that merges with an outlet of the restricted lane, and

removing, from the candidate lane segments, any lane segment that neighbors an inlet of the restricted lane; and

select, from the appropriate nearest lane conflicts, the nearest lane segment into which the vehicle will enter when moving along the reference path to traverse the second road, as the nearest lane conflict.

21. A memory containing programming instructions for inferring a stop line for a vehicle at an intersection, the memory comprising:

a non-transitory computer-readable medium; and

programming instructions stored on the computer-readable medium that, when executed by a processor, will cause the processor to, while the vehicle is moving along a reference path on a first road in a geographic area:

access map data comprising data representing a plurality of roads in the geographic area, wherein the map data segments each of the roads into lane segments;

detect from the accessed map data that a traffic control measure is positioned before an intersection that is ahead of the vehicle in the reference path;

determine whether a stop line for the detected traffic control measure is painted on the first road before or at the intersection;

in response to determining that no stop line is painted on the first road before or at the intersection, perform the following operations:

identifying, as a restricted lane, a lane segment of the first road that (i) includes an entry line proximate a location of the traffic control measure and (ii) extends into the intersection, wherein the entry line at least partially defines a boundary of the restricted lane;

identifying, as a nearest lane conflict, a lane segment of a second road intersecting the restricted lane;

determining whether the entry line of the restricted lane is within a target distance of the lane segment of the second road intersecting the restricted lane;

setting the inferred location of the stop line to a location of the entry line responsive to a determination being made that the entry line is within the target distance of the lane segment of the second road; and

performing the following other operation responsive to a determination being made that the entry line is not within the target distance of the lane segment of the second road:

initializing a location of an inferred stop line as the location of the entry line of the restricted lane;

advancing the location of the inferred stop line a distance away from the vehicle and towards the nearest lane conflict;

iteratively repeating the advancing until the location of the inferred stop line is at the target distance from a nearest boundary of the nearest lane conflict;

setting a final location of the inferred stop line as the location of the inferred stop line at completion of said iteratively repeating;

refining the final location of the inferred stop line based on at least one object perceived in an environment using sensor data, wherein the at least one object excludes any traffic control measure; and

setting the inferred location of the stop line as the refined final location of the inferred stop line.