DETECTING RISKY DRIVING WITH MACHINE VISION

• US 20170061222A1
• Filed: 08/31/2015
• Published: 03/02/2017
• Est. Priority Date: 08/31/2015
• Status: Active Grant

• Current Assignee: Lytx, Inc.
• Original Assignee: Lytx, Inc.
• Inventors: Hoye, Brett, Surpi, Gabriela
• Granted Patent: US 9,996,756 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle
• CPC Class Codes: G06V 20/597 Recognising the driver's state or behaviour, e.g. attention or drowsiness
• CPC Class Codes: G07C 5/008 communicating information to a remotely located station transmission systems for measured values G08C
• CPC Class Codes: G07C 5/06 in graphical form
• CPC Class Codes: G07C 5/0825 using optical means
• CPC Class Codes: G07C 5/0891 Video recorder in combination with video camera

Abstract

A system for driver risk determination using machine vision includes an interface and one or more processors. The interface is to receive one or more video data streams. The one or more processors is/are to determine, whether a driving behavior risk type appears in one of the one or more video data streams; and in the event that the driving behavior risk type appears in the one of the one or more video data streams, indicate the driver behavior risk type.

Claims

1. A system for driver risk determination using machine vision, comprising:

an interface to receive one or more video data streams; and

one or more processors to;

determine, using machine vision, whether a driving behavior risk type appears in one of the one or more video data streams; and

in the event that the driving behavior risk type appears in the one of the one or more video data streams, indicate the driver behavior risk type.

2. The system of claim 1, wherein the processor is further to:

determine an incident severity of driver behavior risk based at least in part on the driver behavior risk type.

3. The system of claim 2, wherein the incident severity of a driver behavior risk type is determined based at least in part on a fraction of time the behavior was sustained.

4. The system of claim 2, wherein the incident severity of driver behavior risk is determined is based at least in part on how many driver behavior risk types co-occur in the incident.

5. The system of claim 2, wherein one processor of the one or more processors is further to:

in the event the incident severity of driver behavior risk is above a driver behavior risk threshold, provide behavior risk type to a server.

6. The system of claim 2, wherein one processor of the one or more processors is further to:

in the event the incident severity of driver behavior risk is above a driver behavior risk threshold, store video data from the one or more of the one or more video data streams.

7. The system of claim 2, wherein one processor of the one or more processors is further to:

in the event the incident severity of driver behavior risk is above a driver behavior risk threshold, provide driver behavior risk metadata and the one or more of the one or more video data streams to a server.

8. The system of claim 7, wherein the server is used to determine a past or a future driver risk based at least in part on the driver behavior risk metadata or the one of the one or more video data streams.

9. The system of claim 8, wherein the past or the future driver risk is based at least in part on a behavior observed during a human review of the one of the one or more video data streams.

10. The system of claim 8, wherein the driver risk is determined based at least in part on a number of driver behavior events of the driver behavior risk type.

11. The system of claim 8, wherein the driver risk is determined based at least in part on a number of driver behavior events of a plurality of driver behavior risk types.

12. The system of claim 8, wherein the driver risk is based at least in part on a number of driver behavior events of the driver behavior risk type occurring during a recent period of time.

13. The system of claim 8, wherein the driver risk is based at least in part on a number of driver behavior events of a plurality of types occurring during recent periods of time.

14. The system of claim 8, wherein the driver risk is based at least in part on a severity of a determined driving behavior risk incidence.

15. The system of claim 8, wherein the driver risk is based at least in part on a severity of a is determined driving behavior risk type.

16. The system of claim 1, wherein one processor of the one or more processors is further to:

in the event a driver risk is above a threshold, indicate a coachable event.

17. The system of claim 1, wherein the one of the one or more video data streams comprises an interior camera and the driver behavior risk type is determined based at least in part on a gaze direction.

18. The system of claim 17, wherein the gaze direction comprises a gaze not directed at the road ahead.

19. The system of claim 1, wherein the one of the one or more video data streams comprises an interior camera and the driver behavior risk type is determined based at least in part on a head orientation.

20. The system of claim 19, wherein the head orientation comprises a direction not pointing at the road ahead.

21. The system of claim 1, wherein the one of the one or more video data streams comprises an interior camera and the driver behavior risk type is determined based at least in part on a body position.

22. The system of claim 1, wherein the one of the one or more video data streams comprises an interior camera and the driver behavior risk type is determined based at least in part on a hand position.

23. The system of claim 22, wherein the hand position comprises a hand position holding a phone.

24. The system of claim 22, wherein the hand position comprises a hand position during eating.

25. The system of claim 1, wherein the one of the one or more video data streams comprises an interior camera and the driver behavior risk type is determined based at least in part on an eyelid position.

26. The system of claim 25 wherein the eyelid position comprises an eyelid position of is closed eyes.

27. The system of claim 1, wherein one of the one or more video data streams comprises an exterior camera and the driver behavior risk type is determined based at least in part on a lane drift.

28. The system of claim 1, wherein one of the one or more video data streams comprises an exterior camera and the driver behavior risk type is determined based at least in part on a lane choice.

29. The system of claim 1, wherein the one of the one or more video data streams comprises an exterior camera and the driver behavior risk type is determined based at least in part on a turn indicator.

30. The system of claim 1, wherein the one of the one or more video data streams comprises an exterior camera and the driver behavior risk type is determined based at least in part on a speed relative to a traffic speed.

31. The system of claim 1, wherein the one of the one or more video data streams comprises an exterior camera and the driver behavior risk type is determined based at least in part on a lane change frequency.

32. The system of claim 1, wherein the one or more video data streams comprises one or more of the following:

a video stream from an interior camera, a video stream from an exterior camera, a video stream from a vehicle camera, or a video stream from an interior camera and a video stream from an exterior camera.

33. A method for driver risk determination using machine vision, comprising:

receiving one or more video data streams;

determining, using one or more processors, whether a driving behavior risk type appears in one of the one or more video data streams; and

in the event that a driving behavior risk type appears in the one of the one or more video data streams, indicating the driver behavior risk type.

34. A computer program product for driver risk determination using machine vision, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:

receiving one or more video data streams;

determining, using machine vision, whether a driving behavior risk type appears in one of the one or more video data streams; and

in the event that a driving behavior risk type appears in the one of the one or more video data streams, indicating the driver behavior risk type.