Methods for Detecting Vehicle Following Distance
Systems, methods, models, and training data for models are discussed, for determining vehicle positioning, and in particular identifying tailgating. Simulated training images showing vehicles following other vehicles, under various conditions, are generated using a virtual environment. Models are trained to determine following distance between two vehicles. Trained models are used in detection of tailgating, based on determined distance between two vehicles. Results of tailgating are output to warn a driver, or to provide a report on driver behavior. Following distance over time is determined, and simplified following distance data is generated for use at a management device.
1 . A method for determining a following distance between a first vehicle and a second vehicle, the method comprising:
accessing an image including a representation of the first vehicle from a perspective of the second vehicle behind the first vehicle, the image further representing a common lane of travel of the first vehicle and the second vehicle;
determining, by at least one processor, a first vertical position in the image representing a bottom of the first vehicle;
accessing, by the at least one processor, a second vertical position in the image representing a static physical distance from the second vehicle;
determining a transformed first vertical position by applying, by the at least one processor, an image transformation matrix to the first vertical position;
determining a transformed second vertical position by applying, by the at least one processor, the image transformation matrix to the second vertical position;
determining an image distance between the transformed first vertical position and the transformed second vertical position;
determining the following distance as a physical distance between the first vehicle and the second vehicle based on the determined image distance and the static physical distance; and
outputting the determined following distance.
2 . The method of claim 1 , wherein the first vertical position represents a bottom boundary of a bounding box which encompasses the first vehicle.
3 . The method of claim 1 , wherein the second vertical position represents a distal end of a hood of the second vehicle as represented in the image.
4 . The method of claim 1 , wherein the image transformation matrix represents a transformation of the image to a transformed image having a fixed relationship between pixel size in the transformed image and physical distance represented by the transformed image.
5 . The method of claim 1 , further comprising:
determining a first boundary of the common lane of travel;
determining a second boundary of the common lane of travel;
determining a vanishing point for the image as a point where the first boundary and the second boundary intersect;
accessing a region of interest in the image, where a bottom edge and a top edge of the region of interest are parallel to a horizontal axis of the image, a left edge of the region of interest extends from first point left of the common lane of travel towards the vanishing point, and a right edge of the region of interest extends from a second point right of the common lane of travel towards the vanishing point;
determining a transformed region of interest where a left edge of the transformed region of interest is parallel to a right edge of the transformed region of interest; and
determining the image transformation matrix as a matrix which when applied to the region of interest, transforms the region of interest to the determined transformed region of interest.
6 . The method of claim 5 , wherein the first point is at a left boundary of the image and the second point is at a right boundary of the image.
7 . The method of claim 5 , wherein the bottom edge of the region of interest is positioned above a hood of the second vehicle as represented in the image, and the top edge of the region of interest is positioned below the vanishing point.
8 . The method of claim 5 , wherein accessing the region of interest in the image comprises receiving an indication of the region of interest as user input via a user input device.
9 . The method of claim 5 , wherein accessing the region of interest in the image comprises generating, by the at least one processor, boundaries of the region of interest.
10 . The method of claim 5 , wherein determining the transformed region of interest comprises determining four corner points of a rectangle corresponding to the transformed region of interest.
11 . The method of claim 5 , further comprising:
determining a ratio of image distance to physical distance for a vertical axis of the image as transformed according to the image transformation matrix, based on a received measurement for a physical distance coaxial to the common lane and an image distance coaxial to the common lane as represented in the transformed region of interest.
12 . The method of claim 5 , further comprising:
determining a ratio of image distance to physical distance for a horizontal axis of the image as transformed according to the image transformation matrix, based on a received measurement for a physical width of the common lane and an image distance width of the common lane as represented in the transformed region of interest.
13 . The method of claim 12 , further comprising:
determining a ratio of image distance to physical distance for a vertical axis of the image as transformed according to the image transformation matrix, based on the ratio of image distance to physical distance for the horizontal axis of the image and based on the image transformation matrix.
14 . The method of claim 1 , wherein the static physical distance is received as a physical measurement between a front of the second vehicle and content represented in the image at the second vertical position.
15 . The method of any one of claims 1 to 14 , further comprising:
determining, by the at least one processor, whether the following distance is within a tailgating distance criteria;
identifying, by the at least one processor, that the second vehicle is not tailgating the first vehicle if the following distance is outside of the tailgating distance criteria; and
identifying, by the at least one processor, that the second vehicle is tailgating the first vehicle if tailgating criteria are met, wherein the tailgating criteria includes the following distance being within the tailgating distance criteria.
16 . The method of claim 15 , further comprising, in response to an identification that the second vehicle is tailgating the first vehicle, outputting a tailgating indication to a driver of the second vehicle.
17 . The method of claim 15 , further comprising, in response to an identification that the second vehicle is tailgating the first vehicle, outputting a tailgating indication to a management server.
18 . The method of any one of claims 1 to 17 , accessing the image comprises capturing the image by at least one image capture device of the system.
19 . The method of any one of claims 1 to 17 , wherein accessing the image comprises receiving, by at least one communication interface of the system communicatively coupled to the at least one processor, the image.
20 . The method of any one of claims 1 to 17 , wherein accessing the image comprises accessing the image as stored in at least one non-transitory processor-readable storage medium.