Large vehicle side lamp interference identification and lane keeping method in dark light environment

Abstract

The invention relates to the field of traffic safety, in particular to a large vehicle side lamp interference identification and lane keeping method in a dark light environment. According to the method, when an obstacle target exists, the moving speed V0 of a highlight feature relative to a self-vehicle is compared with the first relative speed V1 of the obstacle target relative to the self-vehicle to determine whether the highlight feature corresponding to a current road image is a correct lane line or a light spot formed by irradiation of a side lamp of a large vehicle or other slender shape features on a vehicle body; and an accurate lane center position is calculated through lane width priori knowledge to reduce the interference of large vehicles on a lane keeping system in a dark environment, so that the vehicles can safely run in the center of the own lane.

Description

Technical Field

[0001] The present invention relates to the technical field of traffic safety, and in particular to a method for identifying interference from sidelights of large vehicles and maintaining lanes in a dark environment. Background Art

[0002] At night on highways or in low-light conditions like tunnels, large vehicles use side lights to ensure driving safety and illuminate blind spots. These lights are mounted on the sides of the vehicle, shining downward, parallel to the vehicle. If the side lights shine on lane markings, the strong reflection from the road surface can sometimes make it difficult to discern lane markings, as the side lights are typically brighter and ambient light is weaker.

[0003] White or yellow lane markings on dark roads appear brighter than the road surface. Therefore, vision-based autonomous driving systems identify lane markings by extracting thin, high-brightness strips from the video feed of the road ahead. However, in images captured by vehicles in adjacent lanes behind a large vehicle, the area illuminated by sidelights appears as a significantly brighter spot than other areas. If the road surface is highly reflective, the lane-keeping system will struggle to accurately capture the lane markings. If the lane markings of the current lane are not captured, the lane-keeping system will mistake the sidelights on the road or other thin, long features on the large vehicle for the lane markings. Consequently, the vehicle will drift toward the large vehicle to reach the desired lane center position, potentially leading to serious accidents. While existing autonomous driving perception methods based on lidar or millimeter-wave radar can accurately perceive other vehicles on the road, radar deployment is expensive, information density is sparse compared to vision, and equipment maintenance is high. Therefore, cost-effectively obtaining reliable lane centering results and preventing interference from sidelights on the lane-keeping system are crucial.

[0004] In view of this, the present invention is proposed. Summary of the Invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a method for identifying interference from sidelights of large vehicles and maintaining lanes in a dark environment.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] On the one hand, the present invention provides a method for identifying interference from side lights of large vehicles and maintaining lanes in a dark environment.

[0008] Obtaining a road image in front of the vehicle and data about the vehicle's surroundings, respectively: extracting corresponding highlight features based on the road image, and determining whether there is an obstacle in front of the vehicle based on the data about the vehicle's surroundings;

[0009] When an obstacle is present, the system first compares the moving speed (V0) of the highlighted feature relative to the ego vehicle with the first relative speed (V1) of the obstacle to determine whether the highlighted feature in the current road image corresponds to a lane line, a light spot formed by the sidelights of a large vehicle, or a slender feature on the vehicle body. The system then uses prior knowledge of lane width to accurately calculate the lane center position. This reduces interference from large vehicles in the lane keeping system in low-light environments and ensures that the current vehicle safely drives in the center of the ego vehicle's lane.

[0010] Furthermore, the method specifically comprises the following steps:

[0011] Step 1: Use a camera to capture an image of the road in front of the vehicle, extract corresponding highlight features based on the road image, and calculate the moving speed V0 of the highlight features relative to the vehicle; scan the road environment in front of the vehicle using a millimeter-wave radar to determine whether there is an obstacle in front of the vehicle;

[0012] Step 2: If there is an obstacle in front of the ego vehicle, calculate the first relative velocity V1 of the obstacle with respect to the ego vehicle, feed the obstacle signal and the first relative velocity V1 back to the lane keeping system, and execute step 3. If there is no obstacle in front of the ego vehicle, feed the obstacle signal back to the lane keeping system, and determine the highlighted feature corresponding to the current road image as the lane line feature.

[0013] Step 3. Calculate the difference between the first relative moving speed V1 and the moving speed V0 of the highlighted feature relative to the ego vehicle. If V1-V0>10%V1, and V0>V, then determine that the highlighted feature corresponding to the current road image is a lane feature moving at high speed relative to the ego vehicle and is unrelated to the obstacle ahead, where V is a set speed threshold. If V1-V0≤10%V1, then determine that the highlighted feature corresponding to the current road image is a light spot formed by the sidelights of a large vehicle or a slender feature on the vehicle body.

[0014] Step 4: If the highlighted features on both sides of the current road image are lane line features that are moving at high speed relative to the ego vehicle, the distances of the lane lines on both sides relative to the ego vehicle's centerline are weighted averaged to ensure that the ego vehicle stays in the center of the lane.

[0015] Step 5: If the highlighted features on both sides of the current road image are light spots formed by the sidelights of a large vehicle or slender features on the vehicle body, the final lane center position is determined based on the dual-threshold feature enhancement model to ensure that the vehicle stays in the lane center.

[0016] Specifically, in step 1, the highlight feature is extracted as follows: the road image is subjected to perspective transformation to extract the corresponding highlight feature; the perspective transformation includes: using the Otsu algorithm to automatically calculate the distribution pattern of each pixel in the road image to extract the highlight feature in the road image.

[0017] Specifically, in step 1, the camera device includes a vehicle-mounted lens, which is installed on the vehicle windshield to capture the road image in front of the vehicle; the millimeter-wave radar is installed on the top of the vehicle to scan the area in front of the vehicle.

[0018] Furthermore, in step 1, the calculation process of the highlighted feature relative to the moving speed V0 of the vehicle is as follows:

[0019] Project the highlighted features at the positions of two consecutive frames of images into the same image;

[0020] Calculate the vertical displacement x of the highlight feature in pixels. Divide the displacement x by the sampling time to get the moving speed V0 of the highlight feature relative to the vehicle in pixels / second.

[0021] According to the calibration relationship, the unit of relative speed is converted into kilometers per hour.

[0022] Specifically, in step 3, if V1-V0≤10%V1, the highlight feature corresponding to the current road image has a circular shape or a pattern of multiple long and thin strip features arranged in parallel. The circular shape mainly presents an irregular shape, an ellipse or a semicircle.

[0023] Specifically, step 4 includes:

[0024] The center line of the current road image is regarded as the center line of the vehicle, and the distances Xl and Xr of the left and right lane lines relative to the center line of the vehicle are calculated respectively;

[0025] Add the distances Xl and Xr of the left and right lane lines relative to the vehicle's centerline and divide by 2 to get the ideal lane center distance.

[0026] Specifically, step 5 includes: if the left highlight feature corresponding to the current road image is a light spot formed by the sidelight of a large vehicle or a slender shape feature on the vehicle body, establishing a dual-threshold feature enhancement model;

[0027] First, use a low threshold segmentation to obtain the mask of the entire highlighted feature area, and then use a high threshold to segment the thin lane features in the highlighted feature area. When there are multiple thin strip features arranged in parallel in the dual threshold feature enhancement result, the thin strip feature closest to the right is selected as the output;

[0028] Based on the prior knowledge of the lane line and lane width on the right side of the vehicle, a virtual lane line is drawn. The final lane center position is determined by comprehensively weighting the distance between the virtual lane line and the left dual-threshold feature enhancement result relative to the vehicle centerline. The virtual lane line is assigned a first weight w1, and the dual-threshold feature enhancement result is assigned a second weight w2, where w1>w2.

[0029] Similarly, if the highlighted feature on the right side of the current road image is a light spot formed by the sidelight of a large vehicle or a slender shape feature on the vehicle body, a dual-threshold feature enhancement model is established;

[0030] First, use a low threshold segmentation to obtain the mask of the entire highlight feature area, and then use a high threshold to segment the thin lane features in the highlight feature area. When there are multiple thin strip features arranged in parallel in the dual threshold feature enhancement result, the thin strip feature closest to the left is selected as the output;

[0031] Based on the prior knowledge of the lane line and lane width on the left side of the vehicle, a virtual lane line is drawn. The final lane center position is determined by comprehensively weighting the distance between the virtual lane line and the dual-threshold feature enhancement result on the right side relative to the center line of the vehicle. Among them, the virtual lane line is assigned a third weight w3, and the dual-threshold feature enhancement result is assigned a fourth weight w4, and w3>w4.

[0032] On the other hand, the present invention also provides a system based on the method for recognizing sidelights of large vehicles and keeping lanes in a dark environment as described above, including:

[0033] An acquisition module is used to obtain the road image in front of the vehicle and the surrounding environment data of the vehicle;

[0034] The judgment module can determine whether there is an obstacle in front of the vehicle based on the vehicle's surrounding environment data;

[0035] When an obstacle is present, the calculation module first compares the moving speed (V0) of the highlighted feature relative to the vehicle with the first relative speed (V1) of the obstacle relative to the vehicle to determine whether the highlighted feature in the current road image is a lane line feature, a light spot formed by the sidelights of a large vehicle, or a slender feature on the vehicle body. The accurate lane center position is then calculated using prior knowledge of lane width.

[0036] The lane keeping system is used to receive signals indicating the presence or absence of an obstacle target and, based on the calculation results of the calculation module, execute the corresponding movement distance to ensure that the current vehicle drives safely in the center of its own lane.

[0037] Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects:

[0038] 1) When an obstacle is present, this method compares the moving speed (V0) of the highlighted feature relative to the ego vehicle with the obstacle's first relative speed (V1). This method can determine whether the highlighted feature corresponding to the current road image is a correct lane line, a light spot formed by the sidelights of a large vehicle, or other slender features on the vehicle body. This method then uses prior knowledge of lane width to accurately calculate the lane center position. This reduces interference from large vehicles to the lane keeping system in low-light environments, ensuring that the vehicle can safely stay in the center of the ego vehicle's lane.

[0039] 2) The proposed method for detecting interference from large vehicle sidelights and maintaining lanes in low-light environments can cost-effectively distinguish lane lines from the light spots produced by large vehicle sidelights or other slender features on the vehicle body. It also sets virtual lane lines based on prior knowledge of lane widths and ensures the vehicle stays in the center of its lane by weighting the distance between the virtual lane lines and the highlighted features relative to the vehicle's centerline. Compared to existing detection methods, this method offers low cost, excellent real-time performance, and precise prediction, and has promising application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are incorporated in and constitute a part of this specification and, together with the description, serve to explain the principles of the invention.

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0042] Figure 1 A flow chart of the method for identifying interference with sidelights of large vehicles and maintaining lanes in a dark environment provided by the present invention;

[0043] Figure 2 A schematic diagram of calculating the moving speed V0 of a highlight feature relative to the vehicle after perspective transformation of an image provided by the present invention;

[0044] Figure 3 A schematic diagram illustrating the characteristics of the combination of features in the highlight feature provided by the present invention;

[0045] Figure 4 Schematic diagram of the distance between the weighted virtual lane line and the highlighted feature relative to the vehicle centerline provided by the present invention;

[0046] Figure 5 This is a schematic diagram of reference point selection for perspective transformation provided by the present invention. DETAILED DESCRIPTION

[0047] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Instead, they are merely examples consistent with certain aspects of the present invention as detailed in the appended claims.

[0048] The present invention provides a method for identifying interference from sidelights of large vehicles and maintaining lanes in a dark environment. The method obtains a road image in front of the vehicle and data about the vehicle's surrounding environment. The method extracts corresponding highlight features based on the road image and determines whether there is an obstacle in front of the vehicle based on the data about the vehicle's surrounding environment.

[0049] When an obstacle is present, the system first compares the moving speed (V0) of the highlighted feature relative to the ego vehicle with the first relative speed (V1) of the obstacle to determine whether the highlighted feature in the current road image corresponds to a lane line, a light spot formed by the sidelights of a large vehicle, or a slender feature on the vehicle body. The system then uses prior knowledge of lane width to accurately calculate the lane center position. This reduces interference from large vehicles in the lane keeping system in low-light environments and ensures that the current vehicle safely drives in the center of the ego vehicle's lane.

[0050] In order to enable those skilled in the art to better understand the technical solutions of the present invention, the present invention is described in further detail below with reference to the accompanying drawings and embodiments.

[0051] Example

[0052] like Figure 1 As shown, this embodiment provides a method for recognizing sidelights of large vehicles and keeping lanes in a dark environment, which specifically includes the following steps:

[0053] Step 1: Use a camera (such as a vehicle-mounted camera) to capture the road image in front of the vehicle, extract the corresponding highlight features based on the road image, and calculate the moving speed V0 of the highlight features relative to the vehicle (see Figure 2 ); Use millimeter-wave radar to scan the road environment in front of the vehicle to determine whether there are obstacles in front of the vehicle. The detailed process is as follows:

[0054] Step 1a: Mount the vehicle camera on the vehicle windshield and point it forward to capture images of the road ahead of the vehicle.

[0055] Step 1b: Select 4 points based on the lane lines in the road image. The selection process is as follows: Figure 5 , through perspective transformation, the lane lines that intersect at point M are transformed into parallel lane lines;

[0056] Step 1c: Use the Otsu algorithm to automatically calculate the distribution pattern of each pixel in the road image to extract the highlight features in the road image, and calculate the moving speed V0 of the highlight features relative to the vehicle;

[0057] Step 1d: Install the millimeter-wave radar on the top of the vehicle and point it forward to scan the front to determine whether there are any obstacles in front of the vehicle.

[0058] The detailed process of calculating the moving speed V0 of the highlighted feature relative to the vehicle in step 1c is as follows:

[0059] Step 1c1, projecting the highlighted features at the positions of two consecutive frames of images into the same image;

[0060] Step 1c2: Calculate the vertical displacement x of the highlighted feature in pixels. Divide the displacement x by the sampling time to obtain the moving speed V0 of the highlighted feature relative to the vehicle in pixels / second.

[0061] Step 1c3: Convert the unit of relative speed to kilometers per hour according to the calibration relationship.

[0062] Step 2: When the millimeter-wave radar detects an obstacle ahead, it calculates a first relative velocity V1 of the obstacle with respect to the vehicle, feeds the obstacle signal and the first relative velocity V1 back to the lane keeping system, and executes step 3. If there is no obstacle ahead of the vehicle, it feeds the obstacle signal back to the lane keeping system, and determines that the highlighted feature corresponding to the current road image is the lane line feature.

[0063] Step 3: Set the speed threshold V = 30 km / h. When there is an obstacle in front of the vehicle, determine the relationship between the moving speed V0 of the highlighted feature relative to the vehicle and the first relative speed V1 of the obstacle relative to the vehicle:

[0064] If V1-V0>10%V1, and V0>V, then the highlighted feature corresponding to the current road image is considered to be the lane line feature relative to the high-speed movement of the ego vehicle, and has nothing to do with the obstacle target ahead. It should be noted here that the first condition indicates that there is a speed difference between the obstacle target and the ego vehicle, and the second condition indicates that the ego vehicle is in normal operation. When the moving speed V0 is greater than 30km / h, the ego vehicle can be considered to be driving normally. When the ego vehicle speed is less than or equal to 30km / h, it is considered to be low at this time and the interference effect does not need to be considered.

[0065] If V1-V0≤10%V1, it is considered that the highlight feature corresponding to the current road image is moving at a similar speed to the obstacle target ahead. The highlight feature is a light spot formed by the sidelights of a large vehicle or other slender features on the vehicle body. In this case, the highlight feature should have a circular shape (irregular shape, elliptical or semicircular shape) or a pattern of multiple slender strip features arranged in parallel.

[0066] Step 4: If the highlighted features on both sides of the current road image are lane line features that are moving at high speed relative to the ego vehicle, then the distances of the lane lines on both sides relative to the ego vehicle's centerline are weighted averaged to ensure that the ego vehicle stays in the center of the lane. The details are as follows:

[0067] Step 4a: Consider the centerline of the current road image as the centerline of the vehicle, and calculate the distances Xl and Xr of the left and right lane lines relative to the centerline of the vehicle, respectively.

[0068] Step 4b: Add the distances Xl and Xr of the left and right lane lines relative to the vehicle's centerline and divide by 2. The resulting distance is considered the ideal lane center distance.

[0069] Step 4c: Control the vehicle to move toward the ideal lane center distance.

[0070] Step 5: When it is determined in step 3 that the highlighted feature on the left side of the ego-vehicle image is a light spot formed by the sidelight of a large vehicle or other slender features on the vehicle body, a dual-threshold feature enhancement model is established;

[0071] First, use a low threshold segmentation to obtain the mask of the entire highlight feature area, and then use a high threshold to segment the thin lane features in the highlight feature area. When there are multiple thin strip features arranged in parallel in the dual threshold feature enhancement result, only the thin strip feature closest to the right is selected as the output;

[0072] Based on the prior knowledge of the lane line and lane width on the right side of the ego vehicle, a virtual lane line is drawn. The distance between the virtual lane line and the left dual-threshold feature enhancement result relative to the ego vehicle centerline is comprehensively weighted. The virtual lane line is assigned a larger first weight w1, and the dual-threshold feature enhancement result is assigned a second weight w2, where w1>w2. The final lane center result is obtained to keep the ego vehicle in the lane center.

[0073] Similarly, see Figures 3-4 When step 3 determines that the highlighted feature on the right side of the ego-vehicle image is a light spot formed by the sidelights of a large vehicle or other slender features on the vehicle body, the process of determining the final lane center position based on the dual-threshold feature enhancement model is as follows:

[0074] Step 5a: Use the Otsu algorithm to automatically calculate the distribution pattern of each pixel in the road image to extract the highlight features in the road image and obtain a highlight feature mask;

[0075] Step 5b: extract the original image of the highlighted feature part according to the highlight feature mask, and use the Otsu algorithm again to automatically calculate the distribution pattern of each pixel in the original image to extract the slender lane feature in the original image;

[0076] Step 5c: When there are multiple long and thin strip features arranged in parallel in the dual-threshold feature enhancement result, only the long and thin strip feature closest to the right is selected as the lane feature output by horizontally comparing the pixel coordinates;

[0077] Step 5d: Consider the centerline of the current road image as the centerline of the vehicle, and calculate the distances Xl and Xr of the left lane line and the right elongated strip feature relative to the centerline of the vehicle, respectively.

[0078] Step 5e: Based on the prior knowledge of lane width, a virtual lane line is drawn at a distance of 3.75 m from the left lane line. The calculated distance of the virtual lane line from the centerline of the vehicle is 3.75 - X1.

[0079] Step 5f: Calculate the ideal lane center distance L according to the following formula:

[0080]

[0081] In the above formula, w1 and w2 are weights that add up to 1, and w1>w2 to reduce the influence of the slender strip feature on the lane center distance;

[0082] Step 5g: Control the vehicle to move toward the ideal lane center distance.

[0083] In summary, the proposed method for recognizing and keeping lanes in dimly lit environments for large vehicles can cost-effectively distinguish lane lines from the light spots produced by large vehicle sidelights or other slender features on the vehicle body. It also sets virtual lane lines based on prior knowledge of lane widths and ensures the vehicle stays in the center of its lane by weighting the distances between the virtual lane lines and the highlighted features relative to the vehicle's centerline. Compared to existing detection methods, this method offers low cost, excellent real-time performance, and precise prediction, and has promising application prospects.

[0084] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention.

[0085] It should be understood that the present invention is not limited to the above description and that various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A method for identifying interference from large vehicle side lights and maintaining lanes in a dark environment, characterized by: Obtaining a road image in front of the vehicle and data about the vehicle's surroundings, respectively: extracting corresponding highlight features based on the road image, and determining whether there is an obstacle in front of the vehicle based on the data about the vehicle's surroundings; When an obstacle is present, the system first compares the moving speed (V0) of the highlighted feature relative to the ego vehicle with the first relative speed (V1) of the obstacle to determine whether the highlighted feature in the current road image corresponds to a lane line, a light spot formed by the sidelights of a large vehicle, or a slender feature on the vehicle body. The system then uses prior knowledge of lane width to accurately calculate the lane center position. This reduces interference from large vehicles in the lane keeping system in low-light environments and ensures that the current vehicle safely drives in the center of the ego vehicle's lane.

2. The method for identifying sidelight interference and keeping lanes of large vehicles in a dark environment according to claim 1, characterized in that: The specific steps include: Step 1: Use a camera to capture an image of the road in front of the vehicle, extract corresponding highlight features based on the road image, and calculate the moving speed V0 of the highlight features relative to the vehicle; scan the road environment in front of the vehicle using a millimeter-wave radar to determine whether there is an obstacle in front of the vehicle; Step 2: If there is an obstacle in front of the ego vehicle, calculate the first relative velocity V1 of the obstacle with respect to the ego vehicle, feed the obstacle signal and the first relative velocity V1 back to the lane keeping system, and execute step 3. If there is no obstacle in front of the ego vehicle, feed the obstacle signal back to the lane keeping system, and determine the highlighted feature corresponding to the current road image as the lane line feature. Step 3. Calculate the difference between the first relative velocity V1 and the moving velocity V0 of the highlighted feature relative to the ego vehicle. If V1-V0>10%V1, and V0>V, then determine that the highlighted feature corresponding to the current road image is a lane feature moving at high speed relative to the ego vehicle and is unrelated to the obstacle ahead, where V is a set velocity threshold. If V1-V0≤10%V1, then determine that the highlighted feature corresponding to the current road image is a light spot formed by the sidelights of a large vehicle or a slender feature on the vehicle body. Step 4: If the highlighted features on both sides of the current road image are lane line features that are moving at high speed relative to the ego vehicle, the distances of the lane lines on both sides relative to the ego vehicle's centerline are weighted averaged to ensure that the ego vehicle stays in the center of the lane. Step 5: If the highlighted features on both sides of the current road image are light spots formed by the sidelights of a large vehicle or slender features on the vehicle body, the final lane center position is determined based on the dual-threshold feature enhancement model to ensure that the vehicle stays in the lane center.

3. The method for identifying sidelight interference and keeping lanes of large vehicles in a dark environment according to claim 2, characterized in that: In step 1, the highlight features are extracted by performing perspective transformation on the road image to extract the corresponding highlight features; the perspective transformation includes using the Otsu algorithm to automatically calculate the distribution pattern of each pixel in the road image to extract the highlight features in the road image.

4. The method for identifying sidelight interference and keeping lanes of large vehicles in a dark environment according to claim 2, characterized in that: In step 1, the camera device includes a vehicle-mounted lens, which is installed on the vehicle windshield to capture the road image in front of the vehicle; the millimeter-wave radar is installed on the top of the vehicle to scan the area in front of the vehicle.

5. The method for identifying sidelight interference and keeping lanes of large vehicles in a dark environment according to claim 2, characterized in that: In step 1, the calculation process of the highlighted feature relative to the moving speed V0 of the vehicle is as follows: Project the highlighted features at the positions of two consecutive frames of images into the same image; Calculate the vertical displacement x of the highlight feature in pixels. Divide the displacement x by the sampling time to get the moving speed V0 of the highlight feature relative to the vehicle in pixels / second. According to the calibration relationship, the unit of relative speed is converted into kilometers per hour.

6. The method for identifying sidelight interference and keeping lanes of large vehicles in a dark environment according to claim 2, characterized in that: In step 3, if V1-V0≤10%V1, the highlight feature corresponding to the current road image has a circular shape or a pattern of multiple long and thin strip features arranged in parallel.

7. The method for identifying sidelight interference and keeping lanes of large vehicles in a dark environment according to claim 2, characterized in that: Step 4 specifically includes: The center line of the current road image is regarded as the center line of the vehicle, and the distances Xl and Xr of the left and right lane lines relative to the center line of the vehicle are calculated respectively; Add the distances Xl and Xr of the left and right lane lines relative to the vehicle's centerline and divide by 2 to get the ideal lane center distance.

8. The method for identifying sidelight interference and keeping lanes of large vehicles in a dark environment according to claim 2, characterized in that: Step 5 specifically includes: if the left highlight feature corresponding to the current road image is a light spot formed by the sidelight of a large vehicle or a slender shape feature on the vehicle body, establishing a dual-threshold feature enhancement model; First, use a low threshold segmentation to obtain the mask of the entire highlighted feature area, and then use a high threshold to segment the thin lane features in the highlighted feature area. When there are multiple thin strip features arranged in parallel in the dual threshold feature enhancement result, the thin strip feature closest to the right is selected as the output; Based on the prior knowledge of the lane line and lane width on the right side of the vehicle, a virtual lane line is drawn. The final lane center position is determined by comprehensively weighting the distance between the virtual lane line and the left dual-threshold feature enhancement result relative to the vehicle centerline. The virtual lane line is assigned a first weight w1, and the dual-threshold feature enhancement result is assigned a second weight w2, where w1>w2.

9. The method for identifying sidelight interference and keeping lanes of large vehicles in a dark environment according to claim 8, characterized in that: Step 5 specifically includes: First, the Otsu algorithm is used to automatically calculate the distribution pattern of each pixel in the road image to extract the highlight features corresponding to the road image and obtain the highlight feature mask; Then, the original image corresponding to the highlight feature is extracted according to the highlight feature mask. Based on the original image, the distribution pattern of each pixel in the original image is automatically calculated using the Otsu algorithm to extract the slender lane feature in the original image; When there are multiple long and thin strip features arranged in parallel in the dual-threshold feature enhancement result, the long and thin strip feature closest to the right is selected as the lane feature output by comparing the pixel coordinates horizontally; The center line of the current road image is regarded as the center line of the vehicle, and the distances Xl and Xr of the left lane line and the right slender strip feature relative to the center line of the vehicle are calculated respectively; Based on the prior knowledge of lane width, a virtual lane line is drawn at a specified distance L1 from the left lane line. The distance of the virtual lane line relative to the center line of the vehicle is calculated to be L1-Xl; Calculate the ideal lane center distance L using the following formula: In the above formula, w1 and w2 are weights that add up to 1, and w1>w2, so as to reduce the influence of the slender strip feature on the lane center distance.

10. A system based on the method for identifying interference with side lights of large vehicles and maintaining lanes in a dark environment according to any one of claims 1 to 9, characterized in that: include: An acquisition module is used to obtain the road image in front of the vehicle and the surrounding environment data of the vehicle; The judgment module can determine whether there is an obstacle in front of the vehicle based on the vehicle's surrounding environment data; When an obstacle is present, the calculation module first compares the moving speed (V0) of the highlighted feature relative to the vehicle with the first relative speed (V1) of the obstacle relative to the vehicle to determine whether the highlighted feature in the current road image is a lane line feature, a light spot formed by the sidelights of a large vehicle, or a slender feature on the vehicle body. The accurate lane center position is then calculated using prior knowledge of lane width. The lane keeping system is used to receive signals indicating the presence or absence of an obstacle target and, based on the calculation results of the calculation module, execute the corresponding movement distance to ensure that the current vehicle drives safely in the center of its own lane.

Images