CN106874875A - A kind of vehicle-mounted lane detection system and method - Google Patents

Abstract

The invention discloses a vehicle-mounted lane line detection system and method. The system includes a processing unit arranged on a driving vehicle, and an image acquisition unit connected with the processing unit; wherein: the image acquisition unit is used for real-time acquisition of The lane picture is sent to the processing unit; the processing unit is used to obtain the lane line sample picture offline/online, and establish a lane line sample library; and extract the RGB pixel value of the lane line in the lane line sample library, and establish G ‑S color model, based on the G‑S color model, input the lane picture during driving, detect the lane lines in the picture through likelihood calculation and threshold segmentation, and extract the lane lines in the picture through Hough transform. The invention can overcome the influence brought by the environment change or the illumination change, and effectively improve the detection effect of the lane line.

Description

System and method for vehicle lane line detection

technical field

The invention relates to the technical field of computer image processing, in particular to a vehicle lane line detection system and method.

Background technique

In recent years, driver assistance systems and unmanned vehicle technology have received widespread attention from researchers. Lane line detection is a crucial step in assisted driving and unmanned vehicle technology. The current mainstream lane line detection methods include lane line detection based on shape features, lane line detection based on color features, and lane line detection based on model features. The lane lines detected by the lane line detection method based on shape features need to have obvious lane line edges and are greatly affected by noise or other interference. Lane line detection methods based on color features are greatly affected by illumination. Lane line detection based on model features will detect a large number of edges irrelevant to lane lines and cannot distinguish detected lane lines of different colors.

Existing lane line detection methods are greatly affected by environmental interference and illumination changes. Therefore, a technical problem that those skilled in the art need to solve is to propose a new lane line detection method for the shortcomings of the existing lane line detection methods, to make up for its shortcomings of being greatly disturbed by the environment, and to further improve the effect of lane line detection .

Contents of the invention

The technical problem to be solved by the present invention is to provide a vehicle-mounted lane line detection system and method for the defect that the detection effect of the lane line detection method in the prior art is poor, and the detection process is easily affected by environmental changes or illumination changes.

The technical solution adopted by the present invention to solve its technical problems is:

The present invention provides a vehicle-mounted lane line detection system, including a processing unit arranged on a driving vehicle, and an image acquisition unit connected to the processing unit; wherein:

The image acquisition unit is used to collect the lane pictures during the driving of the vehicle in real time and send them to the processing unit;

The processing unit is used to obtain the lane line sample picture, establish the lane line sample library; and extract the RGB pixel value of the lane line in the lane line sample library, establish the G-S color model, and input the lane picture in the driving process on the basis of the G-S color model , the lane lines in the picture are detected by likelihood probability calculation and threshold segmentation, and the lane lines in the picture are extracted by Hough transform.

Further, the system of the present invention also includes a display unit connected to the processing unit, for displaying an operation interface and man-machine interaction.

Further, the system of the present invention also includes a storage unit connected to the processing unit, for storing the lane line sample library data and the lane picture data acquired in real time.

Further, the system of the present invention further includes a power supply unit connected to the processing unit, for supplying power to the system.

Further, the processing unit of the present invention includes an offline lane detection mode and an online lane detection mode, and one of the detection modes is selected through the display unit.

The invention provides a vehicle lane line detection method, comprising the following steps:

S1. Obtain images containing lane line samples offline or online, establish a lane line sample library, and obtain lane pictures in real time during vehicle driving;

S2, extract the RGB pixel value of the lane line in the lane line sample library, and establish the G-S color model;

S3. On the basis of the G-S color model, input the first frame of the lane picture during the driving of the vehicle, and detect the lane lines in the picture through likelihood calculation and threshold segmentation;

S4, for the lane line detected in the picture, extract the lane line in the picture by Hough transform;

S5. Calculate the vanishing point of the lane line according to the detected lane line and divide the skyline, set the lane line area in other frame lane pictures in the vehicle driving process according to the skyline, perform steps S3 and S4 on the lane line area, and detect Exit the lane markings and track the lane markings.

Further, the method for obtaining the lane line sample picture in step S1 of the present invention includes: using the public lane line picture data set; obtaining the lane line picture through network search and screening; and collecting the lane line picture through field experiments.

Further, the method for establishing the G-S color model in step S2 of the present invention is:

Let the G-S color model be a 1×3 matrix, expressed as:

μ=[E(R)E(G)E(B)]

Among them, E(R) represents the mean value of R value in the RGB pixel values of the lane line in the sample library, E(G) represents the mean value of G value, E(B) represents the mean value of B value, and u is the mean value of the color model;

In the color space, the covariance is used instead of the variance, and the calculation method of the covariance is:

Among them, V is the covariance of the color model, which is a 3×3 matrix, and the values of its positive diagonal are COV(R,R), COV(G,G), COV(B,B), each in the matrix The calculation formula for elements is:

By calculating the mean u and covariance V of the color model, the G-S color model of the lane line is constructed by the following formula:

Among them, X is a 1×3 matrix, the physical meaning is the pixel value of each point in the picture, and the formula is simplified to: P(X)=(X-μ) ^T ·inv(V)·(X -μ).

Further, the method for detecting the lane line in the picture in step S3 of the present invention is:

Set the threshold to 0.0026, and divide the picture into suspected lane line pixels and non-lane line pixels; after distinguishing the lane line pixels, the image is binarized by threshold segmentation, and the suspected lane line pixels are assigned white, non-lane line pixels The lane line pixels are assigned black.

Further, in the step S4 of the present invention, the method for extracting the lane line in the picture by Hough transform is:

Perform edge detection on the image;

Determine the size of the Hough transform accumulator and allocate memory according to the image size;

Set the threshold, and clear the points whose accumulated value in the Hough transform accumulator is less than the threshold according to the threshold value, that is, these points do not correspond to a straight line in the image domain;

Find the point with the largest accumulative value in the Hough transform accumulator, record this point and continue to find and record the next point with the largest accumulative value until all the accumulative values in the accumulator are zero, and these recorded points correspond to the detected pictures straight line in

Draw a straight line in the image domain according to the detected points, and extract the straight line as the detected lane line.

The beneficial effects produced by the present invention are: the present invention provides a lane line detection system and method that combines online color features and shape features. Lane line detection: the system and method can overcome the impact of environmental changes or light changes, and effectively improve the lane line detection effect.

Description of drawings

The present invention will be further described below in conjunction with accompanying drawing and embodiment, in the accompanying drawing:

Fig. 1 is a system block diagram of an embodiment of the present invention;

Fig. 2 is a flow chart of the method of the embodiment of the present invention.

detailed description

In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

As shown in Figure 1, the vehicle-mounted lane line detection system of the embodiment of the present invention includes a processing unit arranged on a driving vehicle, and an image acquisition unit connected to the processing unit; wherein:

The image acquisition unit is used to collect the lane pictures during the driving of the vehicle in real time and send them to the processing unit;

The processing unit is used to obtain the lane line sample picture, establish the lane line sample library; and extract the RGB pixel value of the lane line in the lane line sample library, establish the G-S color model, and input the lane picture in the driving process on the basis of the G-S color model , the lane lines in the picture are detected by likelihood probability calculation and threshold segmentation, and the lane lines in the picture are extracted by Hough transform.

The system also includes a display unit connected to the processing unit for displaying an operation interface and man-machine interaction. The system also includes a storage unit connected to the processing unit for storing the lane line sample library data and the lane picture data acquired in real time. The system also includes a power supply unit connected to the processing unit for supplying power to the system. The processing unit includes an offline lane detection mode and an online lane detection mode, and one of the detection modes is selected through the display unit.

The vehicle-mounted lane line detection method of the embodiment of the present invention is used to realize the vehicle-mounted lane line detection system of the embodiment of the present invention, comprising the following steps:

S1. Obtain images containing lane line samples offline or online, establish a lane line sample library, and obtain lane pictures in real time during vehicle driving;

S2, extract the RGB pixel value of the lane line in the lane line sample library, and establish the G-S color model;

S3. On the basis of the G-S color model, input the first frame of the lane picture during the driving of the vehicle, and detect the lane lines in the picture through likelihood calculation and threshold segmentation;

S4, for the lane line detected in the picture, extract the lane line in the picture by Hough transform;

S5. Calculate the vanishing point of the lane line according to the detected lane line and divide the skyline, set the lane line area in other frame lane pictures in the vehicle driving process according to the skyline, perform steps S3 and S4 on the lane line area, and detect Exit the lane markings and track the lane markings.

The method for obtaining the lane line sample pictures in step S1 includes: using public lane line picture data sets; acquiring lane line pictures through network search and screening; and collecting lane line pictures through field experiments.

The method for establishing the G-S color model in step S2 is:

Let the G-S color model be a 1×3 matrix, expressed as:

μ=[E(R)E(G)E(B)]

Among them, E(R) represents the mean value of R value in the RGB pixel values of the lane line in the sample library, E(G) represents the mean value of G value, E(B) represents the mean value of B value, and u is the mean value of the color model;

In the color space, the covariance is used instead of the variance, and the calculation method of the covariance is:

Among them, V is the covariance of the color model, which is a 3×3 matrix, and the values of its positive diagonal are COV(R,R), COV(G,G), COV(B,B), each in the matrix The calculation formula for elements is:

By calculating the mean u and covariance V of the color model, the G-S color model of the lane line is constructed by the following formula:

Among them, X is a 1×3 matrix, the physical meaning is the pixel value of each point in the picture, and the formula is simplified to: P(X)=(X-μ) ^T ·inv(V)·(X -μ).

The method for detecting the lane lines in the picture in step S3 is:

Set the threshold to 0.0026, and divide the picture into suspected lane line pixels and non-lane line pixels; after distinguishing the lane line pixels, the image is binarized by threshold segmentation, and the suspected lane line pixels are assigned white, non-lane line pixels The lane line pixels are assigned black.

In step S4, the method of extracting the lane lines in the picture by Hough transform is:

Perform edge detection on the image;

Determine the size of the Hough transform accumulator and allocate memory according to the image size;

Set the threshold, and clear the points whose accumulated value in the Hough transform accumulator is less than the threshold according to the threshold value, that is, these points do not correspond to a straight line in the image domain;

Find the point with the largest accumulative value in the Hough transform accumulator, record this point and continue to find and record the next point with the largest accumulative value until all the accumulative values in the accumulator are zero, and these recorded points correspond to the detected pictures straight line in

Draw a straight line in the image domain according to the detected points, and extract the straight line as the detected lane line.

As shown in FIG. 2 , in another specific embodiment of the present invention, the system includes an image acquisition unit, a processing unit, a display unit, a storage unit, and a power supply unit. The image processing unit is used to collect current image information, the processing unit is used to process image data, the display unit is used for interface display and human-computer interaction, the storage unit is used to store data, and the power supply unit is used to provide power support for the entire system. After the system is started, there are two working modes, namely offline lane marking detection mode and online lane marking detection mode, and the user can freely choose one of the working modes through the display interface. The methods used in the two working modes are described below.

P1 offline lane line detection includes the following steps:

S1. Establish a lane line sample library by pre-acquiring pictures containing lane line samples. Lane line sample pictures can be obtained in the following three ways: one is through public data sets, such as the KITTI data set, etc., the other is to obtain lane line pictures through network search and screening, and the third is to collect lane line pictures through field experiments.

S2. Extract RGB pixel values of lane lines in the sample library, and establish a G-S color model. The elements of the color model are mean and variance, and the mean of the G-S color model in the present invention is a 1×3 matrix, as shown in formula (1).

μ=[E(R)E(G)E(B)] (1)

Among them, E(R) represents the mean value of R in the RGB pixel values of the lane line in the sample library, E(G) represents the mean value of G value, E(B) represents the mean value of B value, and u is the mean value of the color model.

Since there are three values of R, G, and B in the color space, each variance is not independent of each other, so the covariance is used instead of the variance in the color space, and the calculation method of the covariance is shown in formula (2).

Among them, V is the covariance of the color model, which is a 3×3 matrix, and the values of its positive diagonal are COV(R,R), COV(G,G), COV(B,B), each in the matrix The calculation formula of the elements is shown in formula (3).

Through the calculated color model mean u and covariance V, the lane line G-S color model can be constructed by formula (4).

Among them, X is a 1×3 matrix, and the physical meaning is the pixel value of each point in the picture. When calculating, formula (4) can be simplified to formula (5):

P(X)＝(X-μ) ^T ·inv(V)·(X-μ) (5)

S3. Input the picture data of the current first frame on the basis of the G-S color model, and detect the lane lines in the picture through likelihood calculation and threshold segmentation.

Use the formula (5) to calculate the current first frame of the input image, and distinguish the suspected lane line pixels and non-lane line pixels in the image by setting the threshold T. The threshold T can be determined through experiments. If the threshold is too small, it will cause A large number of non-lane line pixels are detected, and too large will cause some lane line pixels to be judged as non-lane line pixels. After experimental verification, it is generally taken as 0.0026. In order to facilitate subsequent calculations, after distinguishing the lane line pixels, the image is binarized through threshold segmentation, specifically assigning the suspected lane line pixels to white and non-lane line pixels to black.

S4. On this basis, the Hough transform is used to extract the lane lines in the picture. Hough transform is a kind of straight line detection method commonly used, is widely used in lane line detection and extraction, in the present invention, the specific process that carries out Hough transform on the basis of G-S color model detection is:

(1) Perform edge detection on the image.

(2) Determine the size of the Hough transform accumulator according to the image size and allocate memory.

(3) Set the threshold, and clear the points in the Hough transform accumulator whose cumulative value is less than the threshold according to the threshold, that is, these points do not correspond to a straight line in the image domain.

(4) Find the point with the largest accumulated value in the Hough transform accumulator, record this point and continue to search and record the next point with the largest accumulated value until all the accumulated values in the accumulator are zero, and these recorded points correspond to the detection to the straight line in the image.

(5) Draw a straight line in the image domain according to the detected points, and extract the straight line as the detected lane line.

S5. The processing method of subsequent N frames of pictures is as follows:

(1) The vanishing point can be calculated and the skyline can be divided by the detected lane line. In the image, the parallel lane lines will intersect at one point, which is the vanishing point. According to the vanishing point, the skyline can be divided. Above the skyline is the non-lane line area, and below the lane line is the lane line area.

(2) According to the divided skyline, the region of interest, that is, the lane line region, can be set in the image, which can reduce a large amount of calculation.

(3) Perform S3-S4 operations in the set area of interest to quickly detect and track lane lines.

P2 online lane line detection includes the following steps:

S1. On the basis of offline lane detection P1, display the lane line candidate set on the user interface. During this process, the threshold T in P1S3 can be appropriately reduced, and all lane lines can be detected as much as possible for the user to choose.

S2. After the user specifies the correct lane line, extract the pixel value of the correct lane line, and establish an online G-S color model. The process is the same as P1S2.

S3. Input the current first frame of picture data based on the online G-S color model, and detect lane lines in the picture through likelihood calculation and threshold segmentation. The process is the same as P1S3.

S4. On this basis, the Hough transform is used to extract the lane lines in the picture. The process is the same as P1S4.

S5. The processing method of subsequent N frames of pictures is the same as that of P1S5.

It should be understood that those skilled in the art can make improvements or changes based on the above description, and all these improvements and changes should belong to the protection scope of the appended claims of the present invention.

Claims (10)

1. A vehicle-mounted lane line detection system, characterized in that, comprises a processing unit arranged on a running vehicle, and an image acquisition unit connected to the processing unit; wherein: The image acquisition unit is used to collect the lane pictures during the driving of the vehicle in real time and send them to the processing unit; The processing unit is used to obtain the lane line sample picture, establish the lane line sample library; and extract the RGB pixel value of the lane line in the lane line sample library, establish the G-S color model, and input the lane picture in the driving process on the basis of the G-S color model , the lane lines in the picture are detected by likelihood probability calculation and threshold segmentation, and the lane lines in the picture are extracted by Hough transform. 2. The vehicle-mounted lane line detection system according to claim 1, characterized in that the system further comprises a display unit connected to the processing unit for displaying an operation interface and human-computer interaction. 3. The vehicle-mounted lane marking detection system according to claim 1, characterized in that the system further comprises a storage unit connected to the processing unit for storing the lane marking sample library data and the lane picture data acquired in real time. 4. The vehicle-mounted lane line detection system according to claim 1, further comprising a power supply unit connected to the processing unit for supplying power to the system. 5. The vehicle-mounted lane detection system according to claim 2, wherein the processing unit includes an offline lane detection mode and an online lane detection mode, and one of the detection modes is selected through the display unit. 6. A vehicle-mounted lane line detection method, comprising the following steps: S1. Obtain images containing lane line samples offline or online, establish a lane line sample library, and obtain lane pictures in real time during vehicle driving; S2, extract the RGB pixel value of the lane line in the lane line sample library, and establish the G-S color model; S3. On the basis of the G-S color model, input the first frame of the lane picture during the driving of the vehicle, and detect the lane lines in the picture through likelihood calculation and threshold segmentation; S4, for the lane line detected in the picture, extract the lane line in the picture by Hough transform; S5. Calculate the vanishing point of the lane line according to the detected lane line and divide the skyline, set the lane line area in other frame lane pictures in the vehicle driving process according to the skyline, perform steps S3 and S4 on the lane line area, and detect Exit the lane markings and track the lane markings. 7. The vehicle-mounted lane line detection method according to claim 6, wherein the method for obtaining the lane line sample picture in step S1 comprises: through a public lane line picture data set; through network search and screening to obtain the lane line picture; Collect lane line pictures through field experiments. 8. The vehicle-mounted lane line detection method according to claim 6, wherein the method for setting up the G-S color model in step S2 is: Let the G-S color model be a 1×3 matrix, expressed as: μ=[E(R) E(G) E(B)] Among them, E(R) represents the mean value of R value in the RGB pixel values of the lane line in the sample library, E(G) represents the mean value of G value, E(B) represents the mean value of B value, and u is the mean value of the color model; In the color space, the covariance is used instead of the variance, and the calculation method of the covariance is: $\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; =</span>\left[\begin{array}{ccc}\mathrm{<span\; class="notranslate">\; C</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; C</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; C</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; )</span>\\ \mathrm{<span\; class="notranslate">\; C</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; C</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; C</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; )</span>\\ \mathrm{<span\; class="notranslate">\; C</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; C</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; C</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; )</span>\end{array}\right]$ Among them, V is the covariance of the color model, which is a 3×3 matrix, and the values of its positive diagonal are COV(R,R), COV(G,G), COV(B,B), each in the matrix The calculation formula for elements is: By calculating the mean u and covariance V of the color model, the G-S color model of the lane line is constructed by the following formula: $\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; V</span>}<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\frac{{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&mu;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}{\mathrm{<span\; class="notranslate">\; V</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&mu;</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span>$ Among them, X is a 1×3 matrix, the physical meaning is the pixel value of each point in the picture, and the formula is simplified to: P(X)=(X-μ) ^T ·inv(V)·(X -μ). 9. The vehicle-mounted lane line detection method according to claim 6, wherein the method for detecting the lane line in the picture in step S3 is: Set the threshold to 0.0026, and divide the picture into suspected lane line pixels and non-lane line pixels; after distinguishing the lane line pixels, the picture is binarized by threshold segmentation, and the suspected lane line pixels are assigned white, non-lane line pixels The pixels of the lane line are assigned the value of black. 10. vehicle-mounted lane line detection method according to claim 6, is characterized in that, in step S4, the method for extracting the lane line in the picture by Hough transform is: Edge detection on the picture; Determine the size of the Hough transform accumulator and allocate memory according to the image size; Set the threshold, and clear the points whose accumulated value in the Hough transform accumulator is less than the threshold according to the threshold, that is, these points do not correspond to a straight line in the image domain; Find the point with the largest accumulative value in the Hough transform accumulator, record this point and continue to find and record the next point with the largest accumulative value until all the accumulative values in the accumulator are zero, and these recorded points correspond to the detected pictures straight line in Draw a straight line in the image domain according to the detected points, and extract the straight line as the detected lane line.