Driver behavior monitoring
Systems and methods provide, implement, and use using a computer-vision based methods of context-sensitive monitoring and characterization of driver behavior. Additional systems and methods are provided for unsupervised learning of action values, monitoring of a driver's environment, and transmitting visual information from a client to a server.
1. A method for assessing driving behavior, comprising:
detecting, by at least one processor of a computing device in communication with a camera, one or more road markings in a field of view of the camera, wherein the camera is mounted on or in a vehicle;
identifying, by the at least one processor, a type of road marking for a road marking of the detected one or more road markings;
determining, by the at least one processor, that the vehicle crossed over the identified road marking;
determining, by the at least one processor, whether an exception condition applies to the crossing over the identified road marking; and
assessing, by the at least one processor, whether the crossing over the identified road marking was inappropriate based on predefined rules associated with the identified type of road marking and on whether the exception condition applies.
2. The method of claim 1 , wherein determining whether an exception condition applies comprises:
recognizing, by the at least one processor, a construction zone using a scene recognition module; wherein the scene recognition module is part of a visual system.
3. The method of claim 1 , wherein determining whether an exception condition applies comprises:
recognizing, by the at least one processor, construction cones directing vehicles to change lanes.
4. The method of claim 3 , wherein recognizing construction cones directing vehicles to change lanes further comprises recognizing that the construction cones are directing vehicles into a shoulder of the road.
5. The method of claim 1 , wherein determining whether an exception condition applies comprises:
analyzing, by the at least one processor, a crowd-sourced data from other drivers, wherein an applicable exception condition is likely if the crowd-sourced data indicates that the other drivers are crossing over the identified road marking at about a same location.
6. A method for assessing driving behavior, comprising:
detecting, by at least one processor of a computing device in communication with a camera, one or more road markings in a field of view of the camera, wherein the camera is mounted on or in a vehicle;
identifying, by the at least one processor, a type of road marking for a road marking of the detected one or more road markings;
determining, by the at least one processor, that the vehicle crossed over the identified road marking;
identifying, by the at least one processor, a presence of an emergency vehicle;
determining, by the at last one processor, that the vehicle pulled over to a side of a road; and
assessing, by the at least one processor, whether the crossing over the identified road marking was inappropriate based on predefined rules associated with the identified type of road marking and on the determination that the vehicle pulled over to the side of the road in the presence of the emergency vehicle.
7. The method of claim 6 , further comprising:
assessing, by the at least one processor, a responsiveness of a driver based on the determination that the vehicle pulled over to the side of the road in the presence of the emergency vehicle.
8. A method for assessing driving behavior, comprising:
detecting, by at least one processor of a computing device in communication with a camera, one or more road markings in a field of view of the camera, wherein the camera is mounted on or in a vehicle;
identifying, by the at least one processor, a type of road marking for a road marking of the detected one or more road markings;
determining, by the at least one processor, that the vehicle crossed over the identified road marking; and
assessing, by the at least one processor, whether the crossing over the identified road marking was inappropriate based on predefined rules associated with the identified type of road marking, wherein the type of road marking is a solid yellow line, a solid white line, or a pair of solid white lines, and the predefined rules indicate that crossing over a solid yellow line, a solid white line, or a pair of solid white lines is inappropriate.
9. The method of claim 8 , wherein the type of road marking is a pair of solid white lines, and further comprising:
inferring, by the at least one processor, that the vehicle crossed into a carpool lane outside of a designated entry location, based on the determination that the vehicle crossed over the pair of solid white lines.
10. The method of claim 8 , further comprising:
transmitting, by the at least one processor, a notification of the inappropriate crossing to a driver of the vehicle.
11. The method of claim 1 , wherein the camera is a forward-facing camera attached to a windshield of the vehicle.
12. A method for assessing driving behavior, comprising:
detecting, by at least one processor of a computing device in communication with a camera, one or more road markings in a field of view of the camera, wherein the camera is mounted on or in a vehicle;
identifying, by the at least one processor, a type of road marking for a road marking of the detected one or more road markings;
determining, by the at least one processor, that the vehicle crossed over the identified road marking;
detecting, by the at least one processor, whether and when the driver signaled a lane change; and
assessing, by the at least one processor, whether the crossing over the identified road marking was inappropriate based on predefined rules associated with the identified type of road marking.
13. A method for assessing driving behavior, comprising:
detecting, by at least one processor of a computing device in communication with a camera, one or more road markings in a field of view of the camera, wherein the camera is mounted on or in a vehicle;
identifying, by the at least one processor, a type of road marking for a road marking of the detected one or more road markings;
determining, by the at least one processor, that the vehicle crossed over the identified road marking;
determining, by the at least one processor, relative distances and speeds of other vehicles around the vehicle; and
assessing, by the at least one processor, whether the crossing over the identified road marking was inappropriate based on predefined rules associated with the identified type of road marking on the relative distances and speeds of the vehicle and nearby vehicles.
14. The method of claim 13 , further comprising:
determining, by the at least one processor, the rate of closure of cars in adjacent lanes; and
modifying the assessment based on the determined rate of closure.
15. The method of claim 14 , wherein if a driver changes into a lane with a fast-approaching car, the distance threshold for an assessment of the safety of a lane change is greater than it would have been if the approaching car were going about the same speed as the vehicle.
16. The method of claim 1 , further comprising:
detecting, by the at least one processor, freeway rumble ridges using an accelerometer; and
inferring, by the at least one processor, crossing into the shoulder based on the detected rumble ridges.
17. A computer program product comprising:
a non-transitory computer-readable medium having instructions stored thereon, the instructions being executable by one or more processors configured to:
detect one or more road markings in a field of view of a camera, wherein the camera is mounted on or in a vehicle;
identify a type of road marking for a road marking of the detected one or more road markings;
determine that the vehicle crossed over the identified road marking;
assess whether the crossing over the identified road marking was inappropriate based on predefined rules associated with the identified type of road marking; and
transmit a notification of the inappropriate crossing to a driver of the vehicle.