Automatic detection of traffic signal states for driver safety system
View Patent ↗A driver safety system estimates a time that a traffic signal transitions (e.g., from red to green) by obtaining location information from a user device in the vehicle at various times and using a physical model of the movement of the vehicle to estimate when the vehicle began to move.
1. A system for estimating a transition time of a traffic signal, comprising:
a user device, wherein the user device obtains location information of an object at various times, the object being proximate to the traffic signal;
a remote device coupled to the user device comprising:
an interaction processor adapted to receive location information from the user device;
an acceleration processor adapted to determine a movement model of the object at a subset of the various times, wherein the movement model includes a wait time before individual cars start moving after the traffic signal changes from red to green, said wait time increasing for cars that are further from the traffic signal in a given lane; and
a transition time estimation processor, adapted to accept as input the location information and the movement model, and to determine therefrom an estimate of a transition time of the traffic signal.
2. The system of claim 1 , wherein the transition time processor is further adapted to accept as input a time lag estimate from a lag processor, the time lag estimate corresponding to an expected difference between the transition time of the traffic signal and movement thereafter of the object.
3. The system of claim 1 , further comprising a learning processor operatively coupled to the acceleration processor and a database, the learning processor adapted to collect light transition data for storage in the database and to aggregate the light transition data for use by the acceleration processor.
4. The system of claim 1 , further comprising a queue location estimation processor operatively coupled to the user device interaction processor, the queue location estimation processor adapted to estimate a queue location for the vehicle based on the location information, the queue location adapted for application by the transition time estimation processor to adjust the estimate of the transition time.
5. The system of claim 4 , further comprising a lag processor operatively coupled to the queue location estimation processor, the lag processor adapted to take as input the queue location and provide as output a time lag estimate.
6. The system of claim 1 , wherein the movement model includes acceleration.
7. The system of claim 1 , wherein the object is a vehicle.
8. The system of claim 1 , wherein the object is a pedestrian.
9. The system of claim 1 , wherein the subset corresponds to at least one time at which the object as not yet passed the traffic signal and another time at which the object has passed the traffic signal.
10. The system of claim 1 , wherein the transition time of the traffic signal is a time at which the signal transitions from red to green.
11. The system of claim 1 , wherein the transition time of the traffic signal is a time at which the signal transitions from green to red.
12. A computer-implemented method of estimating a transition time of a traffic signal, comprising:
receiving, from a user device, location information of an object at various times, the object being proximate to the traffic signal;
applying, via a processor, a movement model of the movement of the object based on the location information to determine an expected time that the object began to move, wherein the movement model includes a wait time before individual cars start moving after the traffic signal changes from red to green, said wait time increasing for cars that are further from the traffic signal in a given lane; and
estimating the transition time of the traffic signal from the expected time that the object began to move.
13. The computer-implemented method of claim 12 , wherein the movement model includes estimation of acceleration of the object from the location information at the various times.
14. The computer-implemented method of claim 12 , wherein the various times include at least one time at which the vehicle has not yet passed the traffic signal and another time at which the vehicle has passed the traffic signal.
15. The computer-implemented method of claim 12 , further comprising collecting lag data for storage in a database and aggregating the lag data to determine a difference between when the traffic light transitioned and when the object began to move.
16. The computer-implemented method of claim 12 , further comprising estimating a queue location for the object based on the location information and adjusting the expected transition time based on the queue location.
17. The computer-implemented method of claim 12 , wherein the object is a vehicle.
18. The computer-implemented method of claim 12 , wherein the object is a pedestrian.
19. The computer-implemented method of claim 12 , further comprising aggregating movement model parameters of additional objects to tune the movement model.
20. The computer-implemented method of claim 12 , further comprising comparing actual observation of the traffic signal with the estimated transition time to tune the movement model.
21. The computer-implemented method of claim 12 , wherein the transition time of the traffic signal is a time at which the signal transitions from red to green.
22. The computer-implemented method of claim 12 , wherein the transition time of the traffic signal is a time at which the signal transitions from green to red.