CN201825037U - Lane departure alarm device for vehicles on highway - Google Patents

Abstract

The utility model discloses a lane departure warning device for expressway vehicles. The device is composed of the following parts: a visual acquisition and processing system, including a camera, an image acquisition card, and a computer. The terminal is connected to the alarm system; the camera collects image information and stores it in the image acquisition card, and the image acquisition card transmits the image information to the computer, and the computer processes the image information and sends an alarm signal to the alarm system according to the processing results; the alarm system consists of audio and Composed of indicator lights, the input ends of the audio and indicator lights are connected to the output ends of the computer. The utility model complements the information based on area segmentation and the information of edge detection, improves the accuracy of lane boundary extraction, and can robustly detect lanes in various complex road environments, using the center line of the vehicle and the left and right lanes The distance is used as the alarm basis, which can effectively carry out lane departure alarm.

Description

Lane departure warning device for highway vehicles

technical field

The utility model belongs to the field of automobile auxiliary driving research, in particular to a lane departure warning device for expressway vehicles. the

Background technique

With the rapid development of expressways in our country, people pay more and more attention to the active safety of vehicles on expressways. Especially in passenger and cargo transportation, due to driver fatigue or other factors, the vehicle deviates from the lane, and traffic accidents such as rear-end collision, side collision, collision with the middle isolation belt or rushing down the road occur, resulting in loss of goods and casualties. The research and development of vehicle lane departure warning device has become very necessary and urgent. the

The correct extraction of lane information in the vehicle lane departure warning device is the premise and basis for the normal operation of the whole device. At present, most of the currently developed vehicle lane departure warning devices use grayscale images for processing, and the extraction of lane information is easily disturbed by factors such as road shadows, oil pollution, incomplete lane markings, and road damage. Completely and accurately extract lane information. Therefore, it is very necessary to study a lane departure warning device that can effectively reflect lane information and effectively send out vehicles. the

Contents of the invention

In view of the defects or deficiencies in the above-mentioned prior art, the purpose of this utility model is to provide a lane departure warning device for expressway vehicles. Compare and identify the values, extract the difference factor, compare this factor with the lane departure control threshold, and issue a lane departure alarm if the threshold is exceeded. the

In order to achieve the above object, the utility model adopts the following technical solutions:

A lane departure alarm device for expressway vehicles, the device is composed of a vision acquisition and processing system and an alarm system:

The visual acquisition and processing system is used to collect the image information of the vehicle position and direction and detect the lane; the system includes a camera, an image acquisition card, and a computer, the output end of the image acquisition card is connected to the input end of the computer, and the output end of the computer is connected to the Alarm system; the camera collects the image information of the vehicle position and direction in real time, and stores the collected image information into the image acquisition card, and the image acquisition card transmits the image information to the computer in real time, and the computer processes the received image information in real time , and send an alarm signal to the alarm system according to the processing result;

The alarm system consists of an indicator light and a sound. The input ends of the sound and the indicator light are connected to the output end of the computer. The indicator light and the sound are used to receive the alarm signal sent by the computer in real time and send out sound and light alarms respectively according to the alarm signal. the

The image acquisition card is an FPGA image acquisition card. the

The utility model also includes the following other technical features:

The image acquisition card is an FPGA image acquisition card. the

The utility model complements the information based on area segmentation and the information of edge detection, improves the accuracy of lane boundary extraction, and can robustly detect lanes in various complex road environments, using the center line of the vehicle and the left and right lanes The distance is used as the alarm basis, which can effectively carry out lane departure alarm. the

Description of drawings

Fig. 1 is a structural schematic diagram of a vehicle lane departure warning device of the present invention. the

Figure 2 is a schematic diagram of the interface of the system equipment. the

Fig. 3 is a working flow diagram of the alarm device of the present invention. the

Fig. 4 is a flow chart of lane detection during the working process of the alarm device. the

Fig. 5 is a schematic diagram of the geometric relationship of lane offset distances. The labels in the figure respectively represent: 1. camera, 2. image, and 3. left boundary line of the lane. the

Detailed ways

As shown in Figure 1, the alarm device of the present utility model is made up of visual collection and processing system, alarm system:

The visual acquisition and processing system is used to collect the image information of the vehicle position and direction and detect the lane; the system includes a camera, an image acquisition card, and a computer. The output end of the image acquisition card is connected to the input end of the computer, and the output end of the computer is connected to the alarm system; The image information of the vehicle position and direction is collected in real time, and the collected image information is stored in the image acquisition card. The image acquisition card transmits the image information to the computer in real time. The alarm system sends out an alarm signal;

The alarm system consists of an indicator light and a sound. The input ends of the sound and the indicator are connected to the output end of the computer. The indicator light and the sound are used to receive the alarm signals sent by the computer in real time and send out sound and light alarms according to the alarm signals. the

The image acquisition card is an FPGA image acquisition card. The sound and light alarm adopts the vehicle's own sound and instrument panel lights. The computer adopts Lenovo Y460 notebook computer, the processor model is Intel Pentium dual-core P6000, and the memory is 2G. the

In practical application, the camera is installed in the front center of the cab, the camera is connected to the image acquisition card, the computer is connected to the image acquisition card, the computer is connected to the dashboard light and the car audio, and the computer uses the image processing technology brought by MATLAB. software package to process the acquired images. the

As shown in Figure 2, the interface of the system connected to the camera uses the Camera Link interface, and the real-time image information is transmitted to the FPGA image acquisition card at high speed through the Camera Link interface for real-time data processing, and the connection between the acquisition card and the computer is realized through the PCI interface. Communication. the

As shown in Figure 3 and Figure 4, the workflow of the present utility model is as follows:

When a vehicle equipped with a lane departure warning device advances along the lane, the camera will store the image information of the collected road surface features into the image acquisition card in real time, and the image acquisition card will send the image information to the computer in real time, and the computer will process the collected image Processing, the computer uses lane detection to obtain the position and direction information of the lane, compares it with the set value, recognizes it, and extracts the difference factor. When the difference factor exceeds the threshold, the system sends an alarm through sound and light. The specific working process of this device is as follows:

The first step: Lane detection; the lane position is determined through lane detection. This lane detection method gives the spatial position of the lane and the direction angle of the lane, which can be directly applied to the judgment of the relative position of the vehicle and the lane. Lane extraction proceeds through the following steps:

1) Extract the lane boundary template

Use the color information to segment the area to obtain the lane boundary template, that is, use the difference between the brightness and saturation of each pixel in the HSV color space, and the information of brightness and chroma to establish a Gaussian model, perform clustering, and fuse the three components The lane area is determined by the area segmentation results, and finally the lane area is subjected to mathematical morphological operations to obtain the lane boundary extraction template. Its steps are described as follows:

(1) Perform region segmentation on the chroma color components;

Given an m×n image, the result S _H of region segmentation on the chroma color component using formula 1 is:

(Formula 1)

Among them: (i, j) represents the pixel coordinates; H(i, j) represents the chromaticity value of the pixel (i, j); u _H and σ _H represent the mean and variance of the chromatic components, respectively, and u _H and σ _H are respectively Using Formula 2 and Formula 3 to calculate:

$u_h=\frac{1}{\mathrm{mm}}\underset{i=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{\mathrm{no}}{\mathrm{\Σ}}}h(i,j)$ (Formula 2)

${\mathrm{\σ}}_h={\left(\frac{1}{\mathrm{mn}}\underset{i=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{\mathrm{no}}{\mathrm{\Σ}}}{(h(i,j)-u_h)}^2\right)}^{1/2}$ (Formula 3)

In the formula, D is a confidence parameter, and in this embodiment, D=2;

(2) Perform region segmentation on the brightness and color components;

Use formula 4 to perform region segmentation results S _V on the brightness and color components:

(公式4) 

(Formula 4)

Among them: (i, j) represents the pixel coordinates; V(i, j) represents the chromaticity value of the pixel (i, j); u _V and σ _V represent the mean and variance of the luminance component respectively, and u _V and σ _V use Equation 5 and Equation 6 are calculated to get:

$u_V=\frac{1}{\mathrm{mn}}\underset{i=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{\mathrm{no}}{\mathrm{\Σ}}}V(i,j)$ (Formula 5)

${\mathrm{\σ}}_V={\left(\frac{1}{\mathrm{mn}}\underset{i=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{\mathrm{no}}{\mathrm{\Σ}}}{(V(i,j)-u_V)}^2\right)}^{1/2}$ (Formula 6)

In the formula, V(i, j) represents the chromaticity value of the pixel (i, j);

(3) Use the difference information of brightness and saturation to perform region segmentation;

According to the brightness value and saturation value, use formula 7 to calculate the difference C:

C(i, j)＝|V(i, j)-S(i, j)| (Formula 7)

Among them, S(i, j) represents the saturation value of pixel (i, j);

According to Equation 8, the result of region segmentation using the difference information of brightness and saturation S _C :

(公式8) 

(Formula 8)

in: $u_C=\frac{1}{\mathrm{mn}}\underset{i=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{\mathrm{no}}{\mathrm{\Σ}}}C(i,j)$ (Formula 9)

${\mathrm{\σ}}_C={\left(\frac{1}{\mathrm{mn}}\underset{i=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{\mathrm{no}}{\mathrm{\Σ}}}{(C(i,j)-u_C)}^2\right)}^{1/2}$ (Formula 10)

(4) Use the logic "or" operation to fuse S _C and S _V by using formula 11 to obtain the preliminary road area segmentation result L _S :

(公式11) 

(Formula 11)

(5) Use the logical "AND" operation to fuse L _S and _SH by using formula 12 to obtain the road area segmentation result L:

(Formula 12)

(6) scan to obtain the lane template M _L ;

Firstly, scan the region segmentation results from the center to both sides, and from the bottom to the top. Each scan line starts from the center of the image to both sides, and records the first encountered point with a pixel value of 0, and the distance from the center to the point The pixels make up the template M _L of the lane.

(7) use formula 13 to carry out mathematical morphology operation to obtain lane boundary template E ₁ ;

${\mathrm{E.}}_1=(m_L\⊕b)-\left(m_L\mathrm{\Θb}\right)$ (Formula 13)

表示数学形态学膨胀运算； Among them, b=[1] _1×1 represents the structural element; Θ represents the corrosion operation of mathematical morphology;

Indicates the mathematical morphology expansion operation;

2) Detection of road edge points

Use the Sobel operator to perform edge detection on the three color components of chroma, brightness, and saturation, and fuse the edge detection results on the three components to obtain the edge-based lane detection results. The steps are as follows:

(1) The Sobel operator is used to detect the edge on the V, S, and H components respectively, and the edge detection results are recorded as E _V , _ES and E _H .

(2) Use the logic "or" operation to fuse the edge detection results on the three components by using formula 14 to obtain the road edge detection result E ₂ , the formula is expressed as follows:

(公式14) 

(Formula 14)

3) Get lane boundary points

Fusion area segmentation results and edge detection results, that is, use the logical "AND" operation to fuse the lane boundary template E ₁ and edge detection E ₂ obtained from the area segmentation to obtain the lane boundary point E by using formula 15:

(Formula 15)

4) Use Hough transform to determine the lane boundary line

Use Hough transform on the obtained lane boundary points to determine the boundary line of the lane;

Hough transform is a typical method of detecting straight lines. Its main idea is to transform the problem of detecting straight lines into finding peak points in the Hough parameter space. Assuming that the straight lines with unknown boundary positions in the image can be described by the equation:

xcos(θ)+ysin(θ)=r (Formula 16)

Any point (x, y) belonging to the line can be described as a sinusoidal curve in the parameter space (r, θ), and the points on the same line correspond to the same point (r, θ) in the parameter space. By detecting (r , θ) to obtain the (r, θ) point with the maximum value, and then to obtain the parameter equation, that is, to obtain the parameter description of the straight line. the

Step 2: Lane departure alarm

1. Determination of the position of the vehicle relative to the lane

As shown in Figure 5, assuming that the camera is perpendicular to the ground, the control parameter offset distance is determined by formula (17):

$\frac{f}{f+h}=\frac{\mathrm{\Δ}{b}_l}{\mathrm{\Δ}{B}_l}$ (Formula 17)

The focal length of the camera is represented by f (mm), the height of the front end of the lens from the ground is represented by H (mm), and the lateral distance between the imaging endpoint of the left boundary line of the lane on the camera projection plane and the center of the vehicle is represented by Δb _l , where ΔB _l is the image of the left boundary line of the lane The actual lateral offset distance from the centerline of the vehicle on the ground corresponding to the endpoint. k (mm/pixel) represents the relationship ratio between the camera plane and the frame memory resolution (pixel). s _l is the lateral offset distance between the imaging end point of the left boundary line of the lane and the center line of the image. Then using formula 18, the offset distance can be obtained as:

$\mathrm{\Δ}{B}_l=(1+\frac{h}{f})\×k\×{\mathrm{the\; s}}_l$ (Formula 18)

Similarly, the lateral offset distance ΔB _r between the right boundary line of the lane and the center line of the vehicle can be obtained.

2. Lane departure alarm

Judging according to Formula 19, when ΔB _l or ΔB _r is less than a given threshold ΔB, the computer sends an alarm control signal to the indicator light and the sound, and the vehicle and the sound give an audible and visual alarm according to the received alarm control signal.

ΔB _r ≤ ΔB or ΔB _l ≤ ΔB (Equation 19)

Among them, the value of ΔB is related to the width of the vehicle with the alarm device installed. If the width of the vehicle is W, the value of ΔB should be W/2. the

Claims (2)

1. A lane departure alarm device for expressway vehicles, characterized in that the device is made up of a visual collection and processing system, an alarm system: The visual acquisition and processing system is used to collect the image information of the vehicle position and direction and detect the lane; the system includes a camera, an image acquisition card, and a computer, the output end of the image acquisition card is connected to the input end of the computer, and the output end of the computer is connected to the Alarm system; the camera collects the image information of the vehicle position and direction in real time, and stores the collected image information into the image acquisition card, and the image acquisition card transmits the image information to the computer in real time, and the computer processes the received image information in real time , and send an alarm signal to the alarm system according to the processing result; The alarm system consists of an indicator light and a sound. The input ends of the sound and the indicator light are connected to the output end of the computer. The indicator light and the sound are used to receive the alarm signal sent by the computer in real time and send out sound and light alarms respectively according to the alarm signal. 2. highway vehicle lane departure warning device as claimed in claim 1, is characterized in that, described image acquisition card is FPGA image acquisition card.

Images