Lane change warning method and system for the vehicle ahead based on image recognition of wheel steering angle

Through grid laser and image recognition technology, combined with the motion model of car variations, the warning of the front car changing behavior is achieved, solving the problem that the existing technology is difficult to identify and warn with the assistance of the existing technology, and improves the warning of early warning. Accuracy and efficiency.

CN115601993BActive Publication Date: 2025-01-28WUHAN UNIV OF TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202210947874.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-08-09
Publication Date
2025-01-28
Estimated Expiration
2042-08-09

AI Technical Summary

Technical Problem

The prior art is difficult to effectively identify and warn of lane change behaviors in front of the vehicle without lane line assistance, and there are limitations on road conditions and vehicle use models.

Method used

By setting up grid coordinates of grid laser, according to the turning wheel image information of the front car, the grid coordinates related to the relevant points are collected and transformed into the position coordinates under the ground coordinate system. Then, by establishing a motion model of a car variation, the rotating angle interval of the car when the vehicle is changed at different speed of the vehicle, and through the comparison of the actual wheel rotation angle and the range of the trail conditions, it is Remind the driver.

Benefits of technology

The early warning of the front -lane changing behavior was achieved with the auxiliary of the lane -free line, reducing the error caused by the pitch angle of the car, and improving the accuracy and efficiency of the early warning. The incidence of rear -end collisions is reduced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115601993B_ABST
    Figure CN115601993B_ABST
Patent Text Reader

Abstract

The present invention discloses a lane change warning method and system for the preceding vehicle based on image recognition of wheel steering angle. The present invention establishes grid coordinates of a grid laser, collects grid coordinates of relevant points according to the steering wheel image information of the preceding vehicle, and converts them into position coordinates in the geodetic coordinate system through geometric relationships, and calculates the actual steering wheel rotation angle and the preceding vehicle speed; by establishing a motion model of the vehicle lane change, the rotation angle range of the vehicle when changing lanes under different vehicle speed conditions is obtained; by comparing the actual wheel rotation angle with the lane change condition range, the lane change behavior of the driver in front is predicted and the driver is reminded. The present invention can predict the lane change behavior of the preceding vehicle without lane line assistance, shorten the driver's reaction time to brake and slow down the vehicle, and achieve the purpose of reducing the incidence of rear-end collision accidents.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of assisted driving, and in particular to a method and system for warning a preceding vehicle of lane change based on image recognition of wheel steering angles. Background Art

[0002] With the continuous development of the economy, the number of cars in my country has increased year by year, and road traffic safety has been paid more and more attention by people. In the actual driving process, some drivers will make abnormal lane changes due to wrong judgments, which will bring great danger to the following vehicles. At this time, the driving skills and reaction speed of the following car drivers are very high.

[0003] In order to increase driving safety, automobile manufacturers have proposed many implementation methods for the recognition and warning of lane changes of the vehicle in front, such as identifying the distance between the lane line of the vehicle in front and the vehicle itself, detecting the acceleration of the edge offset of the vehicle in front, etc. These methods are not ideal on roads without lane lines or require a large amount of calculation during use, have high system requirements, and are limited in road conditions and vehicle types. Summary of the invention

[0004] The present invention aims to provide a method and system for warning the lane change of a preceding vehicle based on image recognition of wheel steering angles, which can be used in the field of assisted driving and the field of automatic driving, and can at least warn the lane change behavior of the preceding vehicle to a certain extent.

[0005] The present invention is achieved through the following technical solutions:

[0006] The method for warning lane change of a preceding vehicle based on image recognition of wheel steering angles provided by the present invention is characterized in that grid coordinates of a grid laser are established, grid coordinates of relevant points are collected and converted into position coordinates in a geodetic coordinate system through geometric relationships according to the steering wheel image information of the preceding vehicle, and the actual steering wheel rotation angle and the preceding vehicle speed are calculated; a motion model of the automobile lane change is established to obtain the rotation angle range of the automobile when changing lanes under different vehicle speed conditions; and the lane change behavior of the preceding driver is predicted and the driver is reminded by comparing the actual wheel rotation angle with the range of the lane change condition.

[0007] The above-mentioned lane change warning method for the preceding vehicle based on image recognition of wheel steering angle may include the following steps:

[0008] In the first step, a laser grid coordinate system with a default spacing is generated by a variable-distance grid laser. The image of the steering wheel of the vehicle in front is collected and the wheel width is identified. The grid spacing is corrected according to the grid coordinates occupied by the wheel width, and a laser grid coordinate system with a suitable spacing is re-established.

[0009] The second step is to collect the steering wheel image of the front vehicle and identify the front and rear points of the steering wheel at the same height. A pair of front and rear points of the steering wheel at the same height are selected based on the principle of equal height and distance, and the positions of the front and rear points of the steering wheel at the same height in the laser grid coordinate system are calculated.

[0010] The third step is to convert the position coordinates of the front and rear points of the steering wheel at the same height in the laser grid coordinate system into the position coordinates in the geodetic coordinate system through the grid coordinate transformation formula, and calculate the steering wheel rotation angle and the driving speed of the front vehicle through the tangent method and vector synthesis method;

[0011] The fourth step is to establish a three-degree-of-freedom motion model of the front vehicle's lane change process, including the vehicle's forward motion along the x-axis, the vehicle's lateral motion along the y-axis, and the vehicle's pitch motion around the y-axis. The process simulation method is used to calculate the steering wheel's rotation angle range when the front vehicle changes lanes under different vehicle speed conditions;

[0012] The fifth step is to judge the danger level caused by the lane change of the front vehicle by comparing the steering wheel rotation angle of the front vehicle with the steering wheel rotation angle range when changing lanes, and to warn the driver of the front vehicle of the lane change through the alarm unit in different sound and light methods.

[0013] The variable distance grid laser is specifically a square grid laser, and its default grid spacing is 50×50; the grid spacing correction coefficient can be calculated by the grid coordinates occupied by the wheel width and the grid coordinate width occupied by the wheel at the appropriate spacing:

[0014]

[0015] Where: u f is the horizontal coordinate of the front and rear points at the same height of the steering wheel at the front end of the wheel in the default laser grid coordinate system, u r L is the horizontal coordinate of the front and rear points at the same height of the steering wheel at the rear end of the wheel in the default laser grid coordinate system, h is the grid coordinate width occupied by the wheels at appropriate spacing.

[0016] The preceding vehicle lane change warning method based on image recognition of wheel steering angle can collect the deformation of the front and rear suspensions through the suspension height sensor, and calculate the displacement value caused by the pitch angle on the vertical distance between the grid laser and the ground as follows:

[0017]

[0018] Where x1 is the deformation of the front suspension, x2 is the deformation of the rear suspension, L is the projection length of the distance between the front and rear suspensions along the x-axis, and a is the projection length of the distance between the grid laser and the center of mass along the x-axis.

[0019] The preceding vehicle lane change warning method based on image recognition of wheel steering angle converts the acquired grid coordinate data of the target point into the earth coordinate system, specifically:

[0020]

[0021] In the above formula: u is the horizontal coordinate of the front and rear points at the same height of the steering wheel in the laser grid coordinate system; v is the vertical coordinate in the laser grid coordinate system; z0 is the vertical distance between the grid laser and the ground; Δz is the deviation of z0 caused by the pitch angle caused by the driving process of the car; L is the distance between the center point of the grid and the intersection of the projection of the laser grid on the vertical ground when the laser is irradiated at a certain angle to the ground; m is the spacing of the default laser grid generated by vertically irradiating the ground at the same height of the laser; β is the grid spacing correction coefficient, x(u, v) is the horizontal coordinate of the front and rear points at the same height of the steering wheel in the geodetic coordinate system, and y(u, v) is the horizontal coordinate of the front and rear points at the same height of the steering wheel in the geodetic coordinate system.

[0022] The lane change warning method for the preceding vehicle based on image recognition of wheel steering angle is characterized in that when calculating the steering wheel rotation angle of the preceding vehicle, a pair of suitable target points should be selected at the junction of the steering wheel hub and the sidewall. The selected target points should meet the principle of "equal height and far distance":

[0023] (1) The selected pairs of target points should have the same height in the actual coordinate system or be centrally symmetric about the center point of the hub. That is, they must satisfy:

[0024] x f (u, v)-x0(u, v)=x r (u, v)-x0(u, v),

[0025] In the above formula, x f (u, v) is the actual horizontal coordinate of the front end point of the steering wheel at the same height, x r (u, v) is the actual horizontal coordinate of the rear end point of the steering wheel at the same height, and x0(u, v) is the actual horizontal coordinate of the center point of the wheel hub;

[0026] (2) Considering the systematic error of the image recognition method, if the selected front and rear points at the same height of the steering wheel are too close to the center of the wheel hub, the systematic error will further amplify the calculation error of the steering wheel rotation angle. Therefore, the front and rear points at the same height of the steering wheel that are far away from the center of the wheel hub in the x-axis direction should be selected as much as possible. Therefore, the selected coordinate points should satisfy:

[0027]

[0028] In the above formula, x i(u, v) is the actual horizontal coordinate of the front and rear points at the same height of the steering wheel selected by the wheel, x0(u, v) is the actual horizontal coordinate of the center point of the hub, and R is the hub radius of the steering wheel.

[0029] The preceding vehicle lane change warning method based on image recognition of wheel steering angle can be used to calculate the angle of the preceding vehicle steering wheel by the tangent method after selecting several pairs of front and rear points of the steering wheel at the same height that meet the acquisition requirements:

[0030]

[0031] θ i =arctan(α i ),

[0032]

[0033] Where: y if (u, v), y ir (u, v) is the actual ordinate of the points at the same height before and after the steering wheel at the front and rear ends of the wheels, x if (u, v), x ir (u, v) is the actual horizontal coordinate of the front and rear points of the same height of the steering wheel at the front and rear ends of the wheel, α i is the tangent value of the steering wheel angle calculated for the front and rear points at the same height of different steering wheels, θ i is the angle between the wheel direction and the vehicle's driving direction, and n is the logarithm of the front and rear points at the same height of the selected steering wheel.

[0034] The motion model of the car changing lanes can be simplified to a three-degree-of-freedom motion model of the lane changing process of the preceding car, that is, when the driver turns the steering wheel to deflect the steering wheel of the car, the car immediately drives straight along the direction of the wheel rotation to enter the target driving lane. When the car enters the target driving lane, the steering wheel of the car quickly deflects, and the car immediately drives straight along the direction of the lane line. Since the wheel deflection time is short, the displacement generated by the car during the steering wheel deflection process can be ignored. Therefore, the corresponding relationship between the current driving speed of the car and the turning angle is:

[0035]

[0036] Where: v a is the current speed of the vehicle, t is the time required for the vehicle to change lanes, L is the width of the lane line of the highway on which the vehicle is traveling, and θ0 is the steering angle of the steering wheel corresponding to the current speed of the vehicle.

[0037] In actual driving, the time t required for a car to change lanes is about 3 to 5 seconds. According to China's standard highway standards, the width of urban roads is 3.75 meters, so the above formula can be accurately expressed as:

[0038]

[0039] If the operator detects that the wheel rotation angle θ of the front vehicle satisfies the following relationship: θ2≤θ, the sound and light alarm unit will warn the driver of the vehicle in front of the vehicle to change lanes.

[0040] The preceding vehicle lane change warning method based on image recognition of wheel steering angle can provide a warning by comparing the steering angle of the preceding vehicle's steering wheel with the steering wheel steering angle range during lane change and grading the warning range:

[0041] If θ≥θ1, a level 1 alarm is issued;

[0042] like A second level alarm is issued;

[0043] like A Level 3 alert is issued.

[0044] like A Level 4 alert is issued.

[0045] If θ2>θ, Then judge The vehicle in front has no intention of changing lanes and can drive normally;.

[0046] The lane change warning system for a leading vehicle based on image recognition of wheel steering angles provided by the present invention comprises a variable distance grid laser, an image sensor, a computing unit, a danger judgment unit and an acoustic and visual alarm unit, all of which are mounted on the vehicle, wherein: the variable distance grid laser is used to send a default laser grid and a corrected laser grid with a suitable spacing; the image sensor is used to collect images in a laser grid coordinate system in real time and upload them to the computing unit; the computing unit is used to identify the front and rear points at the same height of the steering wheel, obtain the laser grid position coordinates of the front and rear points at the same height of the steering wheel, convert the geodetic coordinates, further calculate the steering wheel rotation angle of the leading vehicle and the steering wheel rotation angle interval of the vehicle changing lanes under the current vehicle speed condition, and upload them to the danger judgment unit; the danger judgment unit compares the above two types of data and judges the lane change trend of the leading vehicle and uploads it to the acoustic and visual alarm unit; the acoustic and visual alarm unit is used to send a warning to the driver that the leading vehicle has a lane change trend.

[0047] Compared with the prior art, the present invention has the following main advantages:

[0048] The present invention can predict the lane-changing behavior of the preceding vehicle without the assistance of lane lines. After obtaining the coordinates of the steering wheel of the preceding vehicle in the laser grid coordinate system, the coordinates of the target point in the earth coordinate system are made more accurate through the optimized coordinate grid transformation formula, which can reduce the error caused by the pitch angle of the vehicle to a certain extent. The "equal height and far" principle is proposed in the selection of target points, which can avoid the misjudgment caused by the special position of the target point to a certain extent. Then, the steering angle is calculated according to the simplified motion model, which can reduce the running time and improve the efficiency to a certain extent. Finally, a warning classification is proposed, and different response modes of the sound and light alarm correspond to different emergency levels, which can shorten the driver's reaction time to brake and slow down the vehicle to a certain extent, so as to achieve the purpose of reducing the incidence of rear-end collision accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a structural schematic diagram of the system of the present invention;

[0050] Figure 2 Flow chart of the system of the present invention.

[0051] In the figure: 1. Variable distance grid laser; 2. Image sensor; 3. Calculation unit; 4. Danger judgment unit; 5. Sound and light alarm unit. DETAILED DESCRIPTION

[0052] The present invention discloses a lane change warning method and system for the preceding vehicle based on image recognition of wheel steering angle. The present invention establishes grid coordinates of a grid laser, collects grid coordinates of relevant points according to the steering wheel image information of the preceding vehicle, and converts them into position coordinates in the geodetic coordinate system through geometric relationships, calculates the actual steering wheel rotation angle and the preceding vehicle speed; establishes a motion model of the vehicle lane change, obtains the rotation angle range of the vehicle when changing lanes under different vehicle speed conditions; and predicts the lane change behavior of the preceding driver and reminds the driver by comparing the actual wheel rotation angle with the range of the lane change condition.

[0053] The present invention will be further described in detail below in conjunction with the accompanying drawings, which are intended to explain the present invention rather than to limit it.

[0054] The lane change warning system for the preceding vehicle based on image recognition of wheel steering angle provided by the present invention is a lane change recognition system for the preceding vehicle. Figure 1 As shown, it includes a variable distance grid laser 1, an image sensor 2, a computing unit 3, a danger judgment unit 4 and an audio-visual alarm unit 5, wherein: the variable distance grid laser is connected to a power supply through a wire; the image sensor is connected to the computing unit through a wire; the computing unit is connected to the danger judgment unit through a wire; the danger judgment unit is connected to the audio-visual alarm unit through a wire.

[0055] The variable distance grid laser may be a grid effect red light laser, which is installed in the middle of the air intake grille and is used to send a default laser grid and a corrected laser grid with a suitable spacing.

[0056] The image sensor may be a camera, which is installed in the middle of the air intake grille, slightly above the installation position of the variable distance grid laser, for real-time acquisition of images in the laser grid coordinate system and uploading the image signal to the computing unit.

[0057] The computing unit may be a microprocessor installed behind the dashboard in the vehicle, and is used to identify the front and rear points at the same height of the steering wheel, obtain the laser grid position coordinates of the front and rear points at the same height of the steering wheel, convert the geodetic coordinates, further calculate the steering wheel rotation angle of the leading vehicle and the steering wheel rotation angle range for the vehicle to change lanes under the current vehicle speed conditions, and upload the calculated data to the danger judgment unit.

[0058] The danger judgment unit may be a microprocessor installed behind the dashboard in the vehicle, and is used to compare the above two types of data to determine the alarm level that should be triggered and upload it to the sound and light alarm unit.

[0059] The sound and light alarm unit can be a buzzer and an LED light, which is installed next to the dashboard of the car and is used to send different degrees of warning to the driver according to different danger levels.

[0060] The preceding vehicle lane change warning system based on image recognition of wheel steering angle provided by the present invention has the following working process:

[0061] Step 1: Generate a laser grid coordinate system with a default spacing through a grid effect red laser, collect the steering wheel image of the front vehicle through a camera, identify the wheel width through a computing unit, calculate the grid spacing correction coefficient according to the laser grid coordinates occupied by the wheel width, and re-establish a laser grid coordinate system with a suitable spacing;

[0062] Step 2: The camera is used to collect the image of the steering wheel of the front vehicle, and the microprocessor is used to identify the front and rear points of the steering wheel at the same height and calculate their position coordinates in the laser grid coordinate system, and then convert them into position coordinates in the geodetic coordinate system through the grid coordinate transformation formula;

[0063] Step three, extract the vehicle speed through the vehicle's internal instrument panel controller LAN bus, and calculate the steering wheel rotation angle and the driving speed of the front vehicle in turn based on the above-obtained position coordinates.

[0064] Step 4, using a process simulation method to calculate the rotation angle range of the steering wheel when the front vehicle changes lanes under different vehicle speed conditions;

[0065] Step 5, comparing the steering wheel rotation angle of the front vehicle with the steering wheel rotation angle range when changing lanes, and judging the danger level of the current vehicle when the front vehicle changes lanes based on the comparison result;

[0066] Step six, using a buzzer and LED lights to warn the driver of the vehicle ahead of you of lane changes in different sound and light ways.

[0067] The present invention provides a lane change warning method for a preceding vehicle based on image recognition of wheel steering angles, such as Figure 2 As shown, the following steps are included:

[0068] When there is a vehicle ahead, the variable distance grid laser 1 generates a laser grid coordinate system with a default grid spacing of 50×50, and the vehicle driving direction is defined as the horizontal coordinate. The instrument has a fixed focal length and is more suitable for use under high-speed road conditions.

[0069] Further, the image of the steering wheel of the target vehicle is collected by a camera;

[0070] Further, identify the horizontal coordinates u of the two target points at the front and rear ends of the wheel image in the laser grid coordinate system f and u r ;

[0071] Further, the grid spacing correction coefficient is calculated:

[0072]

[0073] Among them, L h is the grid coordinate width of the wheel at the appropriate spacing, u f is the horizontal coordinate of the front and rear points at the same height of the steering wheel at the front end of the wheel in the default laser grid coordinate system, u r It is the horizontal coordinate of the front and rear points at the same height of the steering wheel at the rear end of the wheel in the default laser grid coordinate system.

[0074] A laser grid coordinate system with appropriate spacing is established based on the correction coefficient, and the new grid spacing is: 50×β.

[0075] Furthermore, the image of the target steering wheel is re-collected on a laser grid coordinate system with appropriate spacing, and the center and edge of the steering wheel hub in the image are identified. Several pairs of appropriate target points are selected at the junction of the steering wheel hub and the sidewall to calculate the steering wheel rotation angle of the front vehicle. The selected target points should meet the principle of "equal height and far distance":

[0076] (1) The selected pairs of target points should have the same height in the actual coordinate system or be centrally symmetric about the center point of the hub. That is, they must satisfy:

[0077] x f (u, v)-x0(u, v)=xr (u, v)-x0(u, v),

[0078] Where u is the horizontal coordinate of the target point in the laser grid coordinate system; v is the vertical coordinate in the laser grid coordinate system. f (u, v) is the actual horizontal coordinate of the target point at the front end of the wheel, x r (u, v) is the actual horizontal coordinate of the target point at the rear end of the wheel, and x0(u, v) is the actual horizontal coordinate of the center point of the wheel hub.

[0079] (2) Considering the systematic error of the image recognition method, if the selected front and rear points at the same height of the steering wheel are too close to the center of the wheel hub, the systematic error will further amplify the calculation error of the steering wheel rotation angle. Therefore, the front and rear points at the same height of the steering wheel that are far away from the center of the wheel hub in the x-axis direction should be selected as much as possible. Therefore, the selected coordinate points should satisfy:

[0080]

[0081] In the above formula: x i (u, v) is the actual horizontal coordinate of the target point selected by the wheel, x0(u, v) is the actual horizontal coordinate of the center point of the hub, and R is the radius of the steering wheel hub.

[0082] Due to emergency acceleration or emergency braking during driving, the load will be transferred between the front and rear axles, resulting in different deformations of the front and rear suspensions of the vehicle. The vehicle will generate a pitch angle along the y-axis and affect the vertical distance between the grid laser and the ground. The deformation of the front and rear suspensions can be collected by the suspension height sensor, and the displacement value caused by the pitch angle on the vertical distance between the grid laser and the ground can be calculated as follows:

[0083]

[0084] Where: x1 is the deformation of the front suspension, x2 is the deformation of the rear suspension, L is the projection length of the distance between the front and rear suspensions along the x-axis, and a is the projection length of the distance between the grid laser and the center of mass along the x-axis.

[0085] Furthermore, after establishing a laser grid coordinate system with appropriate spacing, the image of the steering wheel of the front vehicle is recaptured and the center of the steering wheel hub and the edge of the hub in the image are identified. The computing unit obtains the laser grid position coordinates of the front and rear points of the same height of the steering wheel. The grid coordinate data of the front and rear points of the same height of the steering wheel are converted to the geodetic coordinate system through the grid coordinate transformation formula:

[0086]

[0087] Wherein: u is the horizontal coordinate of the target point in the laser grid coordinate system; v is the vertical coordinate in the laser grid coordinate system; z0 is the vertical distance between the grid laser and the ground; Δz is the deviation of z0 caused by the pitch angle caused by the vehicle driving; L is the distance between the center point of the grid and the intersection of the projection of the laser grid on the vertical ground when the laser is irradiated at a certain angle to the ground; m is the spacing of the default laser grid generated by the laser irradiating the ground vertically at the same height, and β is the grid spacing correction coefficient.

[0088] Furthermore, the selected target points that meet the collection requirements are used to calculate the steering wheel angle of the car through the tangent method, and the calculation formula is:

[0089]

[0090] θ i =arctan(α i )

[0091]

[0092] Where: y if (u, v), y ir (u, v) is the actual ordinate of the points at the same height before and after the steering wheel at the front and rear ends of the wheels, x if (u, v), x ir (u, v) is the actual horizontal coordinate of the front and rear points of the same height of the steering wheel at the front and rear ends of the wheel, α i is the tangent value of the steering wheel angle calculated for the front and rear points at the same height of different steering wheels, θ i is the angle between the wheel direction and the vehicle's driving direction. n is the logarithm of the front and rear points at the same height of the selected steering wheel.

[0093] Furthermore, the motion model of the car changing lanes is simplified to a three-degree-of-freedom motion model of the lane changing process of the preceding car (including the forward motion of the car along the x-axis, the lateral motion of the car along the y-axis, and the pitch motion around the y-axis), that is, when the driver turns the steering wheel to deflect the steering wheel of the car, the car immediately drives straight along the direction of the wheel rotation to enter the target driving lane. When the car enters the target driving lane, the steering wheel of the car deflects rapidly, and the car immediately drives straight along the direction of the lane line. Since the wheel deflection time is short, the displacement generated by the car during the deflection of the steering wheel can be ignored. Therefore, the corresponding relationship between the current driving speed of the car and the turning angle is:

[0094]

[0095] Where: v a is the current speed of the car, t is the time required for the car to change lanes, L is the width of the lane line of the highway, and θ0 is the steering angle of the steering wheel corresponding to the current speed of the car.

[0096] Furthermore, in the actual driving process, the time t required for a car to change lanes is about 3 to 5 seconds. According to the Chinese standard highway standard, the width of an urban highway is 3.75 meters, and the limit value can be obtained:

[0097]

[0098] Where: θ1 represents the upper limit of the steering wheel rotation angle range when changing lanes, θ2 represents the lower limit of the steering wheel rotation angle range when changing lanes, and v a The current speed of the car.

[0099] Furthermore, in order to help the driver judge the emergency situation more quickly and accurately, the alarm is graded by comparing the steering wheel rotation angle of the front vehicle with the steering wheel rotation angle range when changing lanes:

[0100] If θ≥θ1, level 1 alarm;

[0101] like , level 2 alarm;

[0102] like : 3rd level alarm;

[0103] like : Level 4 alarm;

[0104] If θ2>θ, it is judged that the front vehicle has no tendency to change lanes and can drive normally;

[0105] In the above formula, θ represents the steering wheel angle of the preceding vehicle, θ1 represents the upper limit of the steering wheel rotation angle range when changing lanes, and θ2 represents the lower limit of the steering wheel rotation angle range when changing lanes.

[0106] The sound and light alarm unit sends out an alarm of corresponding level to warn the vehicle ahead of you of changing lanes.

[0107] The present invention has the following innovative features:

[0108] 1. A detection method for determining the lane-changing intention of a leading vehicle by using the steering angle of the leading vehicle's steering wheel.

[0109] 2. (Introducing grid spacing correction factor and height deviation caused by pitch angle) Optimize the coordinate grid transformation formula.

[0110] 3. Use the “equal height and far distance” principle to select the target point.

[0111] 4. Use the tangent method to calculate the steering angle of the front vehicle's wheels.

[0112] 5. Simplify the motion model and solve the wheel rotation angle range.

[0113] 6. Design the warning range classification according to the steering angle of the front vehicle’s wheels.

Claims

1. A lane change warning method for a preceding vehicle based on image recognition of wheel steering angle, characterized in that: By establishing the grid coordinates of the grid laser, based on the steering wheel image information of the front vehicle, the grid coordinates of the collected relevant points are converted into position coordinates in the geodetic coordinate system through geometric relationships, and the actual steering wheel rotation angle and the speed of the front vehicle are calculated; by establishing a motion model of the car changing lanes, the rotation angle range of the car when changing lanes under different vehicle speed conditions is obtained; By comparing the actual wheel rotation angle with the lane change condition range, the lane change behavior of the driver ahead is predicted and the driver is reminded; The method comprises the following steps: The first step is to generate a default spacing laser grid coordinate system through a variable distance grid laser, collect the steering wheel image of the front vehicle and identify the wheel width, correct the grid spacing according to the grid coordinates occupied by the wheel width, and re-establish a laser grid coordinate system with a suitable spacing. The second step is to collect the steering wheel image of the front vehicle and identify the front and rear points of the steering wheel at the same height. A pair of front and rear points of the steering wheel at the same height are selected using the principle of equal height and distance, and the position of the front and rear points of the steering wheel at the same height in the laser grid coordinate system is calculated. The third step is to convert the position coordinates of the front and rear points of the steering wheel at the same height in the laser grid coordinate system into the position coordinates in the geodetic coordinate system through the grid coordinate transformation formula, and calculate the steering wheel rotation angle and the driving speed of the front vehicle through the tangent method and vector synthesis method. The fourth step is to establish a three-degree-of-freedom motion model of the front vehicle's lane change process, including the vehicle's forward motion along the x-axis, the vehicle's lateral motion along the y-axis, and the vehicle's pitch motion around the y-axis. The process simulation method is used to calculate the steering wheel rotation angle range when the front vehicle changes lanes under different vehicle speed conditions. The fifth step is to judge the danger level caused by the lane change of the front vehicle by comparing the steering wheel rotation angle of the front vehicle with the steering wheel rotation angle range when changing lanes, and to warn the driver of the front vehicle of the lane change through the alarm unit in different sound and light methods.

2. The lane change warning method for the preceding vehicle based on image recognition of wheel steering angle according to claim 1 is characterized in that: The variable distance grid laser is specifically a square grid laser, and its default grid spacing is 50×50; the grid spacing correction coefficient can be calculated by the grid coordinates occupied by the wheel width and the grid coordinate width occupied by the wheel at the appropriate spacing: Where: u f is the horizontal coordinate of the front and rear points at the same height of the steering wheel at the front end of the wheel in the default laser grid coordinate system, u r L is the horizontal coordinate of the front and rear points at the same height of the steering wheel at the rear end of the wheel in the default laser grid coordinate system, h is the grid coordinate width occupied by the wheels at appropriate spacing.

3. The lane change warning method for the preceding vehicle based on image recognition of wheel steering angle according to claim 1 is characterized in that: The deformation of the front and rear suspensions is collected by the suspension height sensor, and the displacement value caused by the pitch angle on the vertical distance between the grid laser and the ground is calculated as follows: Where x1 is the deformation of the front suspension, x2 is the deformation of the rear suspension, L is the projection length of the distance between the front and rear suspensions along the x-axis, and a is the projection length of the distance between the grid laser and the center of mass along the x-axis.

4. The lane change warning method for the preceding vehicle based on image recognition of wheel steering angle according to claim 1, characterized in that: The grid coordinate data of the acquired target point is converted into the geodetic coordinate system, specifically: In the above formula: u is the horizontal coordinate of the front and rear points at the same height of the steering wheel in the laser grid coordinate system; v is the vertical coordinate in the laser grid coordinate system; z0 is the vertical distance between the grid laser and the ground; Δz is the deviation of z0 caused by the pitch angle caused by the driving process of the car; L is the distance between the center point of the grid and the intersection of the projection of the laser grid on the vertical ground when the laser is irradiated at a certain angle to the ground; m is the spacing of the default laser grid generated by vertically irradiating the ground at the same height of the laser; β is the grid spacing correction coefficient, x(u, v) is the horizontal coordinate of the front and rear points at the same height of the steering wheel in the geodetic coordinate system, and y(u, v) is the horizontal coordinate of the front and rear points at the same height of the steering wheel in the geodetic coordinate system.

5. The lane change warning method for the preceding vehicle based on image recognition of wheel steering angle according to claim 1, characterized in that: When calculating the steering wheel rotation angle of the front vehicle, a pair of appropriate target points should be selected at the junction of the steering wheel hub and the sidewall. The selected target points should meet the principle of "equal height and far distance": (1) The selected pairs of target points should have the same height in the actual coordinate system or be centrally symmetric about the center point of the hub, that is, they should satisfy: x f (u, v)-x0(u, v)=x r (u,v)-x0(u,v) In the above formula, x f (u, v) is the actual horizontal coordinate of the front end point of the steering wheel at the same height, x r (u, v) is the actual horizontal coordinate of the rear end point of the steering wheel at the same height, and x0(u, v) is the actual horizontal coordinate of the center point of the wheel hub; (2) Considering the systematic error of the image recognition method, if the selected front and rear points at the same height of the steering wheel are too close to the center of the wheel hub, the systematic error will further amplify the calculation error of the steering wheel rotation angle. Therefore, the front and rear points at the same height of the steering wheel that are far away from the center of the wheel hub in the x-axis direction should be selected as much as possible. Therefore, the selected coordinate points should satisfy: In the above formula, x i (u, v) is the actual horizontal coordinate of the front and rear points at the same height of the steering wheel selected by the wheel, x0(u, v) is the actual horizontal coordinate of the center point of the hub, and R is the hub radius of the steering wheel.

6. The lane change warning method for the preceding vehicle based on image recognition of wheel steering angle according to claim 1, characterized in that: After selecting several pairs of front and rear points of the same height of the steering wheel that meet the collection requirements, the angle formula of the steering wheel of the front vehicle is calculated by the tangent method: θ i =arctan(α i ), Where: y if (u, v), y ir (u, v) is the actual ordinate of the points at the same height before and after the steering wheel at the front and rear ends of the wheels, x if (u, v), x ir (u, v) is the actual horizontal coordinate of the front and rear points of the same height of the steering wheel at the front and rear ends of the wheel, α i is the tangent value of the steering wheel angle calculated for the front and rear points at the same height of different steering wheels, θ i is the angle between the wheel direction and the vehicle's driving direction, and n is the logarithm of the front and rear points at the same height of the selected steering wheel.

7. The lane change warning method for the preceding vehicle based on image recognition of wheel steering angle according to claim 1 is characterized in that: The motion model of the vehicle changing lanes can be simplified to a three-degree-of-freedom motion model of the lane changing process of the preceding vehicle, that is, when the driver turns the steering wheel to deflect the steering wheel of the vehicle, the vehicle immediately drives straight along the wheel rotation direction to enter the target driving lane. When the vehicle enters the target driving lane, the steering wheel of the vehicle quickly deflects, and the vehicle immediately drives straight along the direction of the lane line. Since the wheel deflection time is short, the displacement generated by the vehicle during the steering wheel deflection process can be ignored. Therefore, the corresponding relationship between the current driving speed of the vehicle and the turning angle is: Where: v a is the current speed of the vehicle, t is the time required for the vehicle to change lanes, L is the width of the lane line of the highway on which it is traveling, and θ0 is the steering angle of the steering wheel corresponding to the current speed of the vehicle; In actual driving, the time t required for a car to change lanes is 3 to 5 seconds. According to China's standard highway standards, the width of urban roads is 3.75 meters, so the above formula can be accurately expressed as: If the operator detects that the wheel rotation angle θ of the front vehicle satisfies the following relationship: θ2≤θ, the sound and light alarm unit will warn the driver of the vehicle in front of the vehicle to change lanes.

8. The lane change warning method for the preceding vehicle based on image recognition of wheel steering angle according to claim 1, characterized in that: The warning is issued by comparing the steering wheel rotation angle of the vehicle in front with the steering wheel rotation angle range when changing lanes, and the warning range is classified according to the warning range: If θ≥θ1, a level 1 alarm is issued; like A second level alarm is issued; like A Level 3 alert is issued. like A Level 4 alert is issued. If θ2≥θ, it is judged that the front vehicle has no tendency to change lanes and can drive normally; In the above formula, θ represents the steering wheel angle of the preceding vehicle, θ1 represents the upper limit of the steering wheel rotation angle range when changing lanes, and θ2 represents the lower limit of the steering wheel rotation angle range when changing lanes.

9. The lane change warning system for the vehicle ahead based on image recognition of wheel steering angle is characterized by: The system is implemented by the method described in claim 1, and includes a variable distance grid laser, an image sensor, a computing unit, a danger judgment unit and an acoustic and visual alarm unit, all of which are mounted on the vehicle, wherein: the variable distance grid laser is used to send a default laser grid and a corrected laser grid with appropriate spacing; the image sensor is used to collect images in a laser grid coordinate system in real time and upload them to the computing unit; the computing unit is used to identify the front and rear points at the same height of the steering wheel, obtain the laser grid position coordinates of the front and rear points at the same height of the steering wheel, convert the geodetic coordinates, and further calculate the steering wheel rotation angle of the leading vehicle and the steering wheel rotation angle range of the vehicle changing lanes under the current vehicle speed condition and upload them to the danger judgment unit; the danger judgment unit compares the above two types of data and judges the lane changing trend of the leading vehicle and uploads it to the acoustic and visual alarm unit; the acoustic and visual alarm unit is used to send a warning to the driver that the leading vehicle has a lane changing trend.

Citation Information

Patent Citations

  • Intelligent automobile lane change warning system and warning method thereof

    CN102424017A

  • Front vehicle driver unsafe driving behavior detection method and device thereof

    CN104916165A