CN117818606B

CN117818606B - Intersection following method and system based on vehicle track and vehicle

Info

Publication number: CN117818606B
Application number: CN202211201267.8A
Authority: CN
Inventors: 吴凡
Original assignee: Wuhu Bethel Intelligent Driving Co ltd
Current assignee: Wuhu Bethel Intelligent Driving Co ltd
Priority date: 2022-09-29
Filing date: 2022-09-29
Publication date: 2025-10-31
Anticipated expiration: 2042-09-29
Also published as: CN117818606A

Abstract

This invention discloses a method, system, and vehicle for vehicle trajectory following at intersections. The method specifically includes the following steps: S1, acquiring real-time images of the area in front of the vehicle, identifying vehicles traveling in the same direction as target vehicles in the images, and fitting the trajectory of the target vehicles; S2, finding the target vehicle's trajectory with the smallest deviation from the current vehicle's trajectory within the target vehicle's trajectory; S3, when lane lines are detected to disappear, identifying the vehicle closest to the current vehicle in the current lane as a pre-following vehicle, and performing follow detection based on the lateral deviation C0′ between the pre-following vehicle's trajectory and the target vehicle's trajectory. If the pre-following vehicle meets the following conditions, it is then designated as the following vehicle. This invention enables automatic vehicle following at intersections, providing guidance for driving at intersections and improving driving safety at intersections.

Description

Intersection following method and system based on vehicle track and vehicle

Technical Field

The invention belongs to the technical field of vehicle following, and particularly relates to an intersection following method and system based on a vehicle track and a vehicle.

Background

When a vehicle passes through a road junction, the vehicle is guided into a lane opposite to the road junction as the vehicle does not have lane lines and navigation information and needs to follow the movement track of the front vehicle to keep an ADAS transverse control function. However, due to the fact that road traffic conditions are complex, the situation that two lane lines of an intersection cannot be aligned frequently occurs, two ends of the intersection are respectively a straight road and a curved road, a plurality of guide lines are arranged at two ends of the intersection, a front vehicle changes a road in the intersection, and the like, so that the vehicle is hesitant in the intersection, a driving route cannot be determined, the steering wheel swings left and right, the experience of a driver is affected, and potential safety hazards are caused.

Disclosure of Invention

The invention provides an intersection following method based on a vehicle track, which automatically realizes vehicle following at an intersection.

The invention is realized in such a way that a vehicle track-based intersection follows, the method specifically comprises the following steps:

s1, acquiring an image in front of the vehicle in real time, taking the vehicle running in the same direction in the image as a target vehicle, and fitting a motion track of the target vehicle;

s2, searching a target motion track with the minimum deviation from the motion track of the vehicle in the motion track of the target vehicle;

And S3, when the disappearance of the lane line is detected, taking the vehicle closest to the vehicle in the lane as a pre-following vehicle, performing following detection based on the transverse deviation C0' of the pre-following vehicle and the movement track of the target vehicle, and taking the pre-following vehicle as the following vehicle if the pre-following vehicle meets the following condition.

Further, the motion trail fitting method of the target vehicle specifically comprises the following steps:

And acquiring n track point coordinates of each target vehicle nearest to the current moment, fitting the n track point coordinates, and acquiring a motion track of the corresponding target vehicle.

Further, the method for acquiring the target motion trail specifically comprises the following steps:

Extracting a track point sequence I from the motion track of the vehicle, extracting a track point sequence II from the motion track of each target vehicle, wherein the track point sequence I and the track points in the track point sequence II have the same longitudinal coordinate x;

Setting corresponding track point weights based on the longitudinal coordinates x, and calculating transverse weighted deviation delta y _sum of each track point sequence II;

the motion track corresponding to the minimum value deltay _sum is the target motion track.

Further, the method for calculating the lateral weighted deviation Δy _sum is specifically as follows:

Δy[i]=w[i]×|y_obj[i]-y_hvt[i]|;

Wherein Δy [ I ] represents the lateral weighted deviation of the ith track point, Δy _sum represents the lateral weighted deviation of the corresponding track point sequence II, y _obj [ I ] represents the lateral coordinate of the ith track point in the track point sequence I, y _hvt [ I ] represents the lateral coordinate of the ith track point in the corresponding track point sequence II, and w [ I ] represents the weight value of the ith longitudinal coordinate.

Further, the larger the value of the vertical coordinate x is, the smaller the weight value is.

Further, the method for calculating the lateral deviation C0' of the movement track of the pre-following vehicle and the target vehicle specifically comprises the following steps:

acquiring a coordinate (x _cipv,y_cipv) of the current position of the pre-following vehicle, inputting a longitudinal coordinate x _cipv into a target motion track, and acquiring a longitudinal pre-aiming value Lateral deviation of the movement track of the pre-following vehicle from the target vehicle

Further, after the motion trail is determined as the target motion trail, the target vehicle corresponding to the target motion trail is marked,

And when the pre-following vehicle does not meet the following condition, detecting whether all the target vehicles are marked, and if not, executing step S2.

Further, the following conditions are specifically as follows:

(1) The lateral deviation C0' is less than the lateral deviation threshold;

(2) The yaw angle C1 is smaller than the yaw angle threshold;

(3) Half curvature C2 is less than the curvature threshold;

If the pre-following vehicle meets the three conditions, the pre-following vehicle meets the following conditions, otherwise, the pre-following vehicle does not meet the following conditions.

The invention is realized in that an intersection following system based on a vehicle track, the system comprising:

a camera arranged on the vehicle and a processor in communication connection with the camera;

the camera captures an image of the front of the vehicle and sends it to the processor, which selects a following vehicle based on the intersection following method based on the vehicle trajectory described above.

The present invention is achieved by a vehicle having the intersection following system based on a vehicle track as described above integrated therein.

The invention realizes the automatic following of vehicles at the intersection, provides guidance for driving at the intersection, and improves the safety of driving at the intersection.

Drawings

Fig. 1 is a flow chart of an intersection following method based on a vehicle track according to an embodiment of the present invention.

Detailed Description

The following detailed description of the embodiments of the invention, given by way of example only, is presented in the accompanying drawings to aid in a more complete, accurate, and thorough understanding of the inventive concepts and aspects of the invention by those skilled in the art.

Fig. 1 is a flow chart of a vehicle track-based intersection following method according to an embodiment of the present invention, where the method specifically includes the following steps:

s1, acquiring images in front of the vehicle in real time, taking the vehicle running in the same direction in the images as a target vehicle, fitting the motion trail of the target vehicle,

In the running process of the vehicle, images in front of the running of the vehicle are acquired in real time, the vehicle in the images is extracted, the vehicle running in the same direction as the vehicle is marked, and the marked vehicle is the target vehicle.

The vehicle coordinate system defined by the embodiment of the invention takes the center of a rear axle of a vehicle of the vehicle as an origin, the running direction (namely, the longitudinal direction) of the vehicle as an x axis, the width direction (namely, the arrangement direction (transverse direction) of lanes) of the vehicle as a y axis, and coordinates (x, y) of a target vehicle under the vehicle coordinate system are acquired in real time during the running process of the vehicle, wherein the motion trail fitting method of the target vehicle is specifically as follows:

The method comprises the steps of obtaining n groups of track coordinates (x, y) of a target vehicle closest to the current moment, fitting the n groups of track coordinates (x, y), and obtaining a motion track of the corresponding target vehicle, wherein a motion track model is specifically as follows:

y=C0+C1×x+C2×x²+C3×x³

wherein, C0, C1, C2 and C3 respectively represent fitting coefficients.

S2, searching a target motion track with the minimum transverse deviation with the motion track of the vehicle in the unlabeled target vehicles, and marking the target vehicle corresponding to the target motion track in the target vehicles;

In the embodiment of the invention, the method for determining the target motion trail specifically comprises the following steps:

Extracting a track point sequence from the motion track of the vehicle, namely a track point sequence I, extracting a track point sequence from the motion track of each target vehicle, namely a track point sequence II, wherein the track point sequence I and the track points in the track point sequence II have the same longitudinal coordinate x;

In the embodiment of the present invention, the value of the longitudinal coordinate x is taken at intervals within a set longitudinal distance, for example, the set longitudinal distance is 50 meters, and the set interval is 5 meters, so that the longitudinal coordinate arrays of the track point sequence I and the track point sequence II are x= [5,10,15,20,25,30,35,40,45,50], and the weight sequences thereof are w= [1,0.97,0.95,0.91,0.87,0.81,0.75,0.67,0.59,0.51] in sequence.

The corresponding track point weights are set based on the longitudinal coordinates x, and the larger the longitudinal coordinates x are, the farther the distance between the corresponding target vehicle and the vehicle is, so that the smaller the weight value is, the transverse weighted deviation of each track point sequence II is calculated, and the calculation method is specifically as follows:

Δy[i]=w[i]×|yo_bj[i]-y_hvt[i]|;

In the embodiment of the invention, the method for calculating the lateral deviation C0' of the movement track of the pre-following vehicle and the target vehicle is specifically as follows:

based on a camera target screening algorithm (also called CIPV object selection, wherein CIPV is closest in-PATH VEHICLE for short), acquiring the coordinates (x _cipv,y_cipv) of the nearest vehicle (i.e. the pre-following vehicle) and the current position of the pre-following vehicle in the lane, inputting a longitudinal coordinate x _cipv into a target motion track, and acquiring a longitudinal pre-aiming value Lateral deviation of the following vehicle from the target vehicle motion trajectory

Detecting whether the pre-following vehicle meets the following condition, if so, taking the pre-following vehicle as the following vehicle, if not, detecting whether the target vehicle is marked completely, if not, executing step S2, if so, failing to follow the road crossing of the front vehicle, and exiting the following function;

In the embodiment of the invention, the following conditions are specifically as follows:

(1) The lateral deviation C0' is smaller than a lateral deviation threshold, for example 3m;

(2) The yaw angle C1 is less than a yaw angle threshold, for example 0.26rad;

(3) The half curvature C2 is smaller than the curvature threshold, for example 0.002 1/m.

The invention also provides an intersection following system based on the vehicle track, which comprises:

the camera is arranged on the vehicle, the processor is in communication connection with the camera, the camera collects images in front of the vehicle and sends the images to the processor, and the processor selects to follow the vehicle based on the intersection following based on the vehicle track.

The invention also provides a vehicle, and the intersection following system based on the vehicle track is integrated on the vehicle.

While the present invention has been described by way of example, it should be apparent that the practice of the invention is not limited by the foregoing, but rather is intended to cover various insubstantial modifications of the method concepts and teachings of the invention, either as applied to other applications without modification, or as applied directly to other applications, without departing from the scope of the invention.

Claims

1. A method for intersection following based on vehicle trajectory, characterized in that the method specifically includes the following steps:

S1. Real-time acquisition of images in front of the vehicle, identifying vehicles traveling in the same direction in the images as target vehicles, and fitting the motion trajectory of the target vehicles.

S2. Find the target trajectory that deviates the least from the trajectory of your own vehicle in the trajectory of the target vehicle.

S3. When the lane line disappears, the vehicle closest to the vehicle in the lane is taken as the pre-following vehicle. Follow detection is performed based on the lateral deviation C0′ between the pre-following vehicle and the target vehicle's trajectory. If the pre-following vehicle meets the following conditions, it is taken as the following vehicle.

The specific method for calculating the lateral deviation C0′ between the trajectory of the vehicle to be followed and the target vehicle is as follows:

Obtain the coordinates (x _cipv , y _cipv ) of the current position of the vehicle to be followed. Input the longitudinal coordinate x _cipv into the target motion trajectory to obtain the longitudinal pre-aiming value. The lateral deviation between the trajectory of the vehicle to be followed and the trajectory of the target vehicle.

2. The intersection following method based on vehicle trajectory as described in claim 1, characterized in that the target vehicle's trajectory fitting method is as follows:

Obtain the coordinates of the n trajectory points closest to the current time for each target vehicle, fit the coordinates of the n trajectory points, and obtain the motion trajectory of the corresponding target vehicle.

3. The intersection following method based on vehicle trajectory as described in claim 1, characterized in that the method for obtaining the target motion trajectory is as follows:

Extract trajectory point sequence I from the trajectory of this vehicle, and extract trajectory point sequence II from the trajectory of each target vehicle. The trajectory points in trajectory point sequence I and trajectory point sequence II have the same vertical coordinate x.

Based on the vertical coordinate x, set the corresponding trajectory point weights and calculate the lateral weighted deviation Δy _sum of each trajectory point sequence II;

The trajectory corresponding to the minimum value Δy _sum is the target trajectory.

4. The intersection following method based on vehicle trajectory as described in claim 3, characterized in that the calculation method of the lateral weighted deviation _Δysum is as follows:

Δy[i]=w[i]×|y _obj [i]-y _hvt [i]|;

Where Δy[i] represents the lateral weighted deviation of the i-th trajectory point, _Δysum represents the lateral weighted deviation of the corresponding trajectory point sequence II, _yobj [i] represents the lateral coordinate of the i-th trajectory point in trajectory point sequence I, _yhvt [i] represents the lateral coordinate of the i-th trajectory point in the corresponding trajectory point sequence II, and w[i] represents the weight value of the i-th vertical coordinate.

5. The intersection following method based on vehicle trajectory as described in claim 4, characterized in that the larger the value of the longitudinal coordinate x, the smaller its weight value.

6. The intersection following method based on vehicle trajectory as described in claim 1, characterized in that, after the motion trajectory is determined as the target motion trajectory, the target vehicle corresponding to the target motion trajectory is marked.

If the vehicle to be followed does not meet the following conditions, check whether all target vehicles are marked. If the detection result is no, proceed to step S2.

7. The intersection following method based on vehicle trajectory as described in claim 1 or 6, characterized in that the following conditions are specifically as follows:

(1) The lateral deviation C0′ is less than the lateral deviation threshold;

(2) The yaw angle C1 is less than the yaw angle threshold;

(3) Half curvature C2 is less than the curvature threshold;

If the vehicle to be followed meets the above three conditions, then the vehicle to be followed meets the following conditions; otherwise, the vehicle to be followed does not meet the following conditions.

8. A vehicle trajectory-based intersection following system, characterized in that the system comprises:

A camera mounted on the vehicle, and a processor that communicates with the camera;

The camera captures an image of the front of the vehicle and sends it to the processor, which selects a vehicle to follow based on the vehicle trajectory-based intersection following method described in any one of claims 1 to 7.

9. A vehicle, characterized in that the vehicle integrates an intersection following system based on vehicle trajectory as described in claim 8.