Traffic signal mapping and detection
View Patent ↗A system and method provides maps identifying the 3D location of traffic lights. The position, location, and orientation of a traffic light may be automatically extrapolated from two or more images. The maps may then be used to assist robotic vehicles or human drivers to identify the location and status of a traffic signal.
1. A method comprising:
accessing a set of images which are associated with geographic locations proximate to a traffic intersection;
for each particular image of the set of images, identifying, by one or more computing devices, brightly-colored red, yellow, and green objects within the particular image;
generating, by the one or more computers, a set of traffic signal labels based on the identified brightly-colored red, yellow and green objects;
filtering, by the one or more computers, the set of traffic signal labels to remove labels for brightly-colored red, yellow, or green objects that correspond to objects other than a traffic signal light; and
determining, by the one or more computers, a three-dimensional location of a traffic signal based on the filtered set of traffic signal labels and the geographic locations of the images of the set of images.
2. The method of claim 1 , further comprising identifying associated labels within two or more images of the set of images, based on the geographic locations of the two or more images of the set of images.
3. The method of claim 2 , wherein determining the three-dimensional location of the traffic signal is further based on the identified associated labels.
4. The method of claim 1 , generating a map based on the three-dimensional location of the traffic signal.
5. The method of claim 4 , further comprising sending the map to a computer associated with a vehicle.
6. The method of claim 1 , wherein the identified brightly colored red, yellow, and green objects are the size and aspect ratios to correspond to one or more traffic signals.
7. The method of claim 1 , wherein identifying the associated labels is further based on an association distance between labeled objects of the set of images.
8. The method of claim 1 , wherein identifying the associated labels is further based on the physical dimensions of one or more traffic signals.
9. The method of claim 1 , further comprising determining a lane associated with the particular traffic signal based on comparing the determined three-dimensional location of the particular traffic signal to a map of lanes through the traffic intersection.
10. A system comprising:
memory storing a set of images which are associated with geographic locations proximate to a traffic intersection;
one or more computers configured to:
for each particular image of the set of images, identify brightly-colored red, yellow, and green objects within the particular image;
generate a set of traffic signal labels based on the identified brightly-colored red, yellow and green objects;
filter the set of traffic signal labels to remove labels for brightly-colored red, yellow, or green objects that correspond to objects other than a traffic signal light; and
determine a three-dimensional location of a traffic signal based on the filtered set of traffic signal labels and the geographic locations of the images of the set of images.
11. The system of claim 10 , wherein the one or more computers are further configured to identify associated labels within two or more images of the set of images, based on the geographic locations of the two or more images of the set of images.
12. The system of claim 10 , wherein the one or more computers are further configured to determine the three-dimensional location of the traffic signal further based on the identified associated labels.
13. The system of claim 10 , wherein the one or more computers are further configured to generate a map based on the three-dimensional location of the traffic signal.
14. The system of claim 13 , further comprising sending the map to a computer associated with a vehicle.
15. The system of claim 10 , wherein the identified brightly colored red, yellow, and green objects are the size and aspect ratios to correspond to one or more traffic signals.
16. The system of claim 10 , wherein identifying the associated labels is based on an association distance between labeled objects of the two or more selected images.
17. The system of claim 10 , wherein identifying the associated labels is based on the physical dimensions of a traffic signal.
18. The system of claim 10 , the one or more computers are further configured to determine a lane associated with the particular traffic signal based on comparing the determined three-dimensional location of the particular traffic signal to a map of lanes through the intersection.
19. A non-transitory tangible computer-readable storage medium on which computer readable instructions of a program are stored, the instructions, when executed by a processor, cause the processor to perform a method, the method comprising:
accessing a set of images which are associated with geographic locations proximate to a traffic intersection;
for each particular image of the set of images, identifying brightly-colored red, yellow, and green objects within the particular image;
generating a set of traffic signal labels based on the identified brightly-colored red, yellow and green objects;
filtering the set of traffic signal labels to remove labels for brightly-colored red, yellow, or green objects that correspond to objects other than a traffic signal light; and
determining a three-dimensional location of a traffic signal based on the filtered set of traffic signal labels and the geographic locations of the images of the set of images.
20. The medium of claim 19 , wherein the method further comprises identifying associated labels within two or more images of the set of images, based on the geographic locations of the two or more images of the set of images.