A vehicle intelligent following method

Abstract

The application provides a vehicle intelligent following method, wherein, in step S1, an image obtained by a visual camera at the front of the vehicle is taken as image information to be analyzed; in step S2, whether there is a vehicle in the front lane and the adjacent lane is analyzed according to the image information to be analyzed; if there is a vehicle in the front lane, a historical driving track of the front vehicle is obtained as a safe driving track of the vehicle through a tracking strategy; if there is no vehicle in the front lane and there is a vehicle in the adjacent lane, an estimated driving track is obtained as the safe driving track of the vehicle through a preprocessing strategy; in step S3, a vehicle model providing the safe driving track is recognized according to the image information to be analyzed, wheel size values of the vehicle are called according to the vehicle model, and a wheel rolling track area of the vehicle is obtained through an expansion strategy according to the wheel size values and the safe driving track; the application has the advantage that a following track which obeys traffic rules and has no collision risk can be planned according to the driving track of the front vehicle.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving, and more specifically to a method and approach for intelligent vehicle following. Background Technology

[0002] Autonomous driving refers to a highly centralized control method in which the work performed by the vehicle driver is fully automated. Autonomous driving methods have functions such as automatic vehicle wake-up and start-up and hibernation, automatic parking, automatic driving, automatic door opening and closing, and automatic fault recovery. They also have multiple operating modes such as normal operation, degraded operation, and operation interruption.

[0003] Currently, overcoming the technical challenge of autonomous driving is a key focus. Autonomous driving includes vehicle centering, lane detection, maintaining a safe distance from the vehicle in front, and intelligent following. However, current autonomous driving technology is still immature, and road conditions are complex. For example, in some areas, lane lines may be blurred or absent, and there may be other vehicles ahead. When autonomous driving passes through an intersection, it relies on real-time detection and judgment of lane lines to determine the vehicle's trajectory. However, if the autonomous vehicle fails to detect lane lines when passing through an intersection, it is difficult to both comply with traffic rules and safely pass through the intersection. Therefore, it is necessary to solve this technical challenge to improve the completeness and safety of autonomous driving. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide a vehicle intelligent following method that can predict and plan a following trajectory that both complies with traffic rules and avoids collision risks based on the driving trajectory of the vehicle in front.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A vehicle intelligent following method includes the following steps:

[0007] Step S1: Obtain the image captured by the vision camera at the front of the vehicle as the image information to be analyzed;

[0008] Step S2: Analyze whether there are vehicles in the current lane and adjacent lanes ahead based on the image information to be analyzed. If there are vehicles in the current lane, the historical driving trajectory of the preceding vehicle is obtained through the tracking strategy as the safe driving trajectory of the current vehicle. If there are no vehicles in the current lane but there are vehicles in the adjacent lane, the estimated driving trajectory is obtained through the preprocessing strategy as the safe driving trajectory of the current vehicle.

[0009] Step S3: Identify the vehicle model that provides the safe driving trajectory based on the image information to be analyzed, retrieve the wheel size value of the vehicle based on the vehicle model, and obtain the wheel rolling trajectory area of ​​the vehicle through an expansion strategy based on the wheel size value and the safe driving trajectory.

[0010] Furthermore, the tracking strategy includes constructing a planar coordinate system with the center of the front of the vehicle as the origin. The horizontal coordinate in the planar coordinate system reflects the lateral displacement of the vehicle and the vertical coordinate reflects the longitudinal displacement of the vehicle. The center point of the rear of the preceding vehicle is calibrated in the planar coordinate system as the target point based on the image information to be analyzed. The target points at different times are connected to obtain the historical driving trajectory of the preceding vehicle.

[0011] Furthermore, the preprocessing strategy includes calibrating the position of the center point of the rear of the preceding vehicle in the adjacent lane in the plane coordinate system as an auxiliary tracking point based on the image information to be analyzed, connecting the auxiliary target points at different times to obtain the estimated driving trajectory of the adjacent vehicles, analyzing the lane width value based on the image information to be analyzed, and then translating the estimated driving trajectory in the plane coordinate system with the lane width value to obtain a safe driving trajectory.

[0012] Furthermore, the preprocessing strategy also includes a preprocessing sub-strategy, which includes extracting the starting points of the lane lines on both sides of the lane the vehicle is about to enter as prediction points from the image information to be analyzed; marking the prediction points in a plane coordinate system and indexing the center point of the two prediction points as the ideal point of the vehicle; obtaining the switching point of the center point of the vehicle's front in the plane coordinate system when the vehicle travels straight until it completely leaves the current lane; connecting the origin and the switching point to obtain the straight trajectory; calculating the turning trajectory by using a trajectory planning function formula with the switching point and the ideal point of the vehicle; and connecting the straight trajectory and the turning trajectory to obtain the safe driving trajectory.

[0013] Furthermore, the expansion strategy includes obtaining the vehicle model, retrieving the vehicle's width and wheel size values ​​based on the vehicle model, comparing the vehicle's wheel size values ​​with the wheel size values ​​of vehicles providing safe driving trajectories to obtain first ratio information, comparing the vehicle's width value with the width value of vehicles providing safe driving trajectories to obtain second ratio information, combining the first and second ratio information to obtain comprehensive change information, and then calculating the wheel rolling trajectory area by indexing the comprehensive change information based on the safe driving trajectory.

[0014] Furthermore, step S2 also includes step S21, which presets a curvature reference interval. Based on the historical driving trajectory or the estimated driving trajectory, it is determined whether the curvature in the trajectory is greater than the curvature reference interval. If any curvature is greater than the curvature reference interval, the starting point of that trajectory segment is taken as the departure starting point. The departure starting point includes the initial point where the preceding vehicle leaves the current safe driving trajectory. The departure starting point and the ideal point of this vehicle are used to calculate the subsequent trajectory through a trajectory formula.

[0015] Furthermore, step S4 is included, where a safe distance value is preset. If there is a vehicle in the adjacent lane, the estimated driving trajectory and the safe driving trajectory are extended respectively. It is determined whether the two extended trajectory lines intersect. If they intersect, the intersection point is marked as the collision point in the plane coordinate system. Based on the collision point, the position of the vehicle and the position of the adjacent vehicle, the distance between the vehicle and the adjacent vehicle is calculated to be equal to the safe distance value. The current position of the vehicle is recorded as the undetermined point. The subsequent trajectory is obtained through a self-learning strategy based on the undetermined point.

[0016] Furthermore, the self-learning strategy includes comparing the size value of the vehicle inserted into the lane with the size value of the initially tracked vehicle based on the image information to be analyzed. If the size value of the vehicle inserted into the lane is less than the size value of the initially tracked vehicle, the historical driving trajectory of the initially tracked vehicle is used as the subsequent trajectory. If the size value of the vehicle inserted into the lane is greater than or equal to the size value of the initially tracked vehicle, the vehicle inserted into the lane is used as the tracking target, and the historical driving trajectory of the tracking target is obtained as the subsequent trajectory.

[0017] Furthermore, step S4 also includes step S41, which presets a following distance value, and determines the distance between the vehicle and the initially tracked vehicle based on the image information to be analyzed as the actual distance value. If the actual distance value is greater than the following distance value, the vehicle stops at the pending point. If the actual distance value is less than or equal to the following distance value, the vehicle issues a siren and continues to use the historical driving trajectory of the initially tracked vehicle as the subsequent trajectory.

[0018] The beneficial effects of this invention are as follows: By acquiring images in front of the vehicle, the situation of vehicles in the current lane or adjacent lanes is judged and selected as the tracking objects of the vehicle. The driving trajectory of the tracking objects is planned as the safe driving trajectory of the vehicle. It can adapt to autonomous driving in scenarios such as intersections or diagonal intersections without lane lines. Furthermore, by acquiring and analyzing images in real time, the vehicle selects whether to plan a safe driving trajectory in different scenarios based on whether the corresponding tracking target changes, so as to ensure that the vehicle can pass through the road segment safely while complying with traffic rules. Attached Figure Description

[0019] Figure 1 This is a flowchart of the steps of the present invention;

[0020] Figure 2 This is the first scene diagram of the present invention;

[0021] Figure 3 This is the second scene diagram of the present invention;

[0022] Figure 4 This is a scenario diagram illustrating the extended strategy of the present invention;

[0023] Figure 5 This is the third scene diagram in this invention;

[0024] Figure 6 This is the fourth scene diagram in this invention. Detailed Implementation

[0025] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Identical components are denoted by the same reference numerals. It should be noted that the terms "front," "rear," "left," "right," "upper," and "lower" used in the following description refer to directions in the accompanying drawings, and the terms "bottom surface," "top surface," "inner," and "outer" refer to directions toward or away from the geometric center of a specific component, respectively.

[0026] Current autonomous driving technology is still immature, and road conditions are complex. For example, in areas where lane markings are unclear or absent, there may be vehicles ahead. Another example is when an autonomous vehicle passes through an intersection. Because autonomous driving relies on real-time lane marking detection to determine the vehicle's trajectory, it may fail to detect lane markings when passing through an intersection, making it difficult to both obey traffic rules and safely navigate such a section. Therefore, it is necessary to address this technical challenge to improve the completeness and safety of autonomous driving. This invention designs a vehicle intelligent following method, such as... Figure 1 and Figure 2 As shown, it includes the following steps:

[0027] Step S1: Obtain the image taken by the vision camera at the front of the vehicle as the image information to be analyzed; the image information to be analyzed contains multiple photos at different times, and the photos contain the vehicle situation in the lane in front of the vehicle and the adjacent lane. The image to be analyzed in this invention contains two lanes in the same direction.

[0028] Step S2: Analyze whether there are vehicles in the current lane and adjacent lanes ahead based on the image information to be analyzed. If there are vehicles in the current lane, the historical driving trajectory of the preceding vehicle is obtained through the tracking strategy as the safe driving trajectory of the current vehicle. If there are no vehicles in the current lane but there are vehicles in the adjacent lane, the estimated driving trajectory is obtained through the preprocessing strategy as the safe driving trajectory of the current vehicle. The result after analysis in the image to be analyzed is that vehicles in the current lane have a higher priority as tracking targets than vehicles in adjacent lanes. The safe driving trajectory in step S2 is the movement trajectory of the vehicle's center point. However, in order to improve the accuracy of tracking the same trajectory, it is necessary to track the wheel trajectory.

[0029] Step S3: Identify the vehicle model that provides the safe driving trajectory based on the image information to be analyzed. Retrieve the wheel size value of the vehicle based on the vehicle model. Obtain the wheel rolling trajectory area of ​​the vehicle through an expansion strategy based on the wheel size value and the safe driving trajectory. By acquiring images in front of the vehicle, determine the vehicle situation in the current lane or adjacent lanes and select them as the tracking objects of the vehicle. Plan the driving trajectory of the tracking objects as the safe driving trajectory of the vehicle. This can adapt to autonomous driving scenarios such as intersections or diagonal intersections without lane lines, so as to ensure that the vehicle can pass through the road segment safely while complying with traffic rules.

[0030] like Figure 2 In the first scenario shown, there is a trackable vehicle in front of this vehicle. The tracking strategy includes constructing a planar coordinate system with the center of the front of this vehicle as the origin. The horizontal axis of the planar coordinate system reflects the lateral displacement of the vehicle, and the vertical axis reflects the longitudinal displacement of the vehicle. The center point of the rear of the preceding vehicle is marked in the planar coordinate system as the target point based on the image information to be analyzed. The historical driving trajectory of the preceding vehicle is obtained by connecting the target points at different times. Assuming the road segment is a crossroads, the preceding vehicle will drive straight to the lane to be entered. The historical driving trajectory of the preceding vehicle will be a straight line, and the corresponding target points will all be on a straight line. If the entrance to the lane to be entered in the road segment is set at an angle, the preceding vehicle needs to turn to enter the lane. Therefore, the preceding vehicle may drive in a straight line for a distance and then turn to enter the lane to be entered. By taking the closely spaced target points, the historical driving trajectory of the preceding vehicle can be determined based on the target points.

[0031] This situation occurs when there are no vehicles in the current lane, and the vehicle is the lead vehicle in that lane, while there are vehicles in the adjacent lane. The preprocessing strategy involves using the image information to determine the position of the center point of the rear of the leading vehicle in the adjacent lane in the planar coordinate system as an auxiliary tracking point. First, the center point of the parking space of the adjacent vehicle is obtained as an auxiliary tracking point. The auxiliary target points at different times are connected to obtain the estimated driving trajectory of the adjacent vehicle. For example, the position of the auxiliary target point at the 1st second of the timer, the position of the auxiliary target point at the 2nd second of the timer, and so on. The lane width value is obtained by analyzing the image information. Then, the safe driving trajectory is obtained by translating the estimated driving trajectory in the planar coordinate system according to the lane width value. This trajectory planning is called the transfer method.

[0032] like Figure 3 In the second scenario shown, there are no vehicles in the current lane, and this vehicle is the lead vehicle in that lane. There are vehicles in adjacent lanes, but these vehicles are either parallel to this vehicle or traveling too slowly, or there are no vehicles in adjacent lanes. In all cases, the preprocessing sub-strategy is used. This preprocessing sub-strategy involves extracting the starting points of the lane lines on both sides of the lane the vehicle is about to enter from the image information to be analyzed, using these as prediction points. These prediction points are then marked in a planar coordinate system, as shown by points K' and K'" in the figure. The center point of these two prediction points is indexed as the ideal position for this vehicle, as shown by point K in the figure. The vehicle then travels in a straight line until it completely leaves the current lane. The center point of the vehicle's front is switched at a point in the planar coordinate system. Connecting the origin and the switching point yields the straight-line trajectory. The switching point and the vehicle's ideal point are then calculated using a trajectory planning function to obtain the turning trajectory. Connecting the straight-line trajectory and the turning trajectory yields the safe driving trajectory. In the second scenario, the entrance to the lane to be entered in the road segment is angled. When the vehicle enters the lane, it needs to turn. This invention first drives the vehicle in a straight line until the vehicle body is completely out of the lane, and then plans the turning curve based on the point. The trajectory planning function involves longitudinal distance and the vehicle's speed.

[0033] like Figure 4 As shown, the system contains a vehicle database with different vehicle model information. Each model corresponds to vehicle size information, such as length, height, width, wheel width and size, etc. The expansion strategy includes obtaining the vehicle model, retrieving the vehicle's width and wheel size values ​​based on the model, comparing the vehicle's wheel size with the wheel size values ​​of vehicles providing safe driving trajectories to obtain first ratio information, comparing the vehicle's width with the width values ​​of vehicles providing safe driving trajectories to obtain second ratio information, combining the first and second ratio information to obtain comprehensive change information, and then calculating the wheel rolling trajectory area by indexing the comprehensive change information based on the safe driving trajectory. For example... Figure 4In the first scenario, car B provides a safe driving trajectory, and car A is the vehicle itself. Analysis shows that the wheel width of car A is the same as that of car B, but car A's width is greater than car B's. Therefore, the analysis indicates that car B's width is f and car A's width is F. The second ratio is f / F. Based on the safe driving trajectory formed by car B, the rolling trajectory area of ​​car A's wheels can be calculated using this ratio. In the second scenario, car B1 provides a safe driving trajectory, and car A1 is the vehicle itself. Analysis shows that car A1's wheel width is greater than that of car B1, but they are the same. Therefore, the analysis indicates that car B1's wheel width is d and car A1's wheel width is D. The first ratio is d / D. Based on the safe driving trajectory formed by car B1, the rolling trajectory area of ​​car A1's wheels can be calculated using this ratio.

[0034] Step S2 also includes step S21, which is based on a preset curvature reference range. It determines whether the curvature in the trajectory is greater than the curvature reference range based on the historical or estimated trajectory. If any curvature is greater than the curvature reference range, the starting point of the trajectory segment is taken as the departure starting point. The departure starting point includes the initial point where the preceding vehicle leaves the current safe driving trajectory. The departure starting point and the ideal point of the current vehicle are calculated using a trajectory formula to obtain the subsequent trajectory.

[0035] For example: Figure 5 In the third scenario shown, the vehicle initially tracks the car in front within its lane. However, after tracking for a certain distance, the car suddenly turns and enters an adjacent lane for some reason. Because changing lanes requires observing more of the surrounding environment, the autonomous driving system does not choose to continue tracking. Instead, the vehicle automatically sets a curvature reference range (0-k). When the vehicle's historical trajectory suddenly deviates to the right, the curvature of each segment of the trajectory is obtained. This curvature is then compared with the curvature reference range. If the curvature of a segment in the historical trajectory is greater than k, it indicates that the car in front has turned significantly. Combined with visual feedback, the system can determine at which point the car turned and changed lanes. As shown in the diagram, points G1 and G2 deviate from the starting point. The first half of the historical trajectory, S1, is retained for the vehicle to continue driving. The vehicle then uses points G1 and G2 as new starting points, obtains points K1 and K2, and plans the subsequent trajectory S2. Figure 4 The trajectory in the later stage is a straight line trajectory. If G1 and K1 are not aligned, and G2 and K2 are not aligned, then the trajectory planning function formula needs to be used to calculate and plan the trajectory so that G1 and K1 are aligned and G2 and K2 are aligned.

[0036] The process also includes step S4, which has a preset safe distance value. If there is a vehicle in an adjacent lane, the estimated driving trajectory and the safe driving trajectory are extended respectively. It is determined whether the two extended trajectories intersect. If they intersect, the intersection point is marked as the collision point in the plane coordinate system. Based on the collision point, the position of the current vehicle, and the position of the adjacent vehicle, the distance between the current vehicle and the adjacent vehicle is calculated. When the distance between the current vehicle and the adjacent vehicle is equal to the safe distance value, the current position of the current vehicle is recorded as the undetermined point. The subsequent trajectory is obtained based on the undetermined point through a self-learning strategy. The self-learning strategy includes comparing the size value of the vehicle inserted into the current lane with the size value of the initially tracked vehicle based on the image information to be analyzed. If the size value of the vehicle inserted into the current lane is smaller than the initial tracking size value, the subsequent trajectory is determined. The initial vehicle size value is used as the historical driving trajectory of the initially tracked vehicle as the subsequent trajectory. If the size value of a vehicle inserted into the current lane is greater than or equal to the size value of the initially tracked vehicle, the vehicle inserted into the current lane is used as the tracking target, and the historical driving trajectory of the tracking target is obtained as the subsequent trajectory. Step S4 also includes step S41, where a following distance value is preset. The distance between the current vehicle and the initially tracked vehicle is determined based on the image information to be analyzed as the actual distance value. If the actual distance value is greater than the following distance value, the current vehicle stops at the pending point. If the actual distance value is less than or equal to the following distance value, the current vehicle issues a siren and continues to use the historical driving trajectory of the initially tracked vehicle as the subsequent trajectory S4.

[0037] For example: Figure 6 In the fourth scenario shown, there are many vehicles in both lanes, and they are queuing to pass. The vehicle has planned its wheel rolling trajectory area and is driving along the planned wheel rolling trajectory area. However, because the autonomous driving system will maintain a safe distance from the vehicle in front, there will be a gap. The adjacent vehicle C will cut in directly because of the gap. Therefore, the system will automatically judge the trajectory of the adjacent vehicle C that is about to cut in. Specifically, it will extend the two trajectories and judge whether they will intersect. If they intersect, it will use visual feedback to verify whether the adjacent vehicle's movement is cutting in. If it is determined that the adjacent vehicle is cutting in, it will determine whether the distance between the two vehicles, the vehicle A and the adjacent vehicle C, is at the safe distance value h after a certain period of time. That is, the vehicle A should be at the pending point H' and the adjacent vehicle should be at the position C'. Then, based on whether the distance between the vehicle A and C' is sufficient for the vehicle C to cut in, and based on the size of the vehicle C and C', it will analyze whether the vehicle A should continue to use the historical trajectory of C' or use the historical trajectory of C after cutting in.

[0038] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for intelligent vehicle following, characterized in that: Includes the following steps: Step S1: Obtain the image captured by the vision camera at the front of the vehicle as the image information to be analyzed; Step S2: Analyze whether there are vehicles in the current lane and adjacent lanes ahead based on the image information to be analyzed. If there are vehicles in the current lane, the historical driving trajectory of the preceding vehicle is obtained through the tracking strategy as the safe driving trajectory of the current vehicle. If there are no vehicles in the current lane but there are vehicles in the adjacent lane, the estimated driving trajectory is obtained through the preprocessing strategy as the safe driving trajectory of the current vehicle. Step S3: Identify the vehicle model that provides the safe driving trajectory based on the image information to be analyzed; retrieve the wheel size value of the vehicle based on the vehicle model; and obtain the wheel rolling trajectory area of ​​the vehicle through an expansion strategy based on the wheel size value and the safe driving trajectory. The tracking strategy includes constructing a planar coordinate system with the center of the front of the vehicle as the origin. The horizontal coordinate in the planar coordinate system reflects the lateral displacement of the vehicle and the vertical coordinate reflects the longitudinal displacement of the vehicle. The center point of the rear of the preceding vehicle is calibrated in the planar coordinate system as the target point based on the image information to be analyzed. The target points at different times are connected to obtain the historical driving trajectory of the preceding vehicle. The preprocessing strategy includes: calibrating the position of the center point of the rear of the preceding vehicle in the adjacent lane in the plane coordinate system according to the image information to be analyzed as an auxiliary tracking point; connecting the auxiliary target points at different times to obtain the estimated driving trajectory of the adjacent vehicle; analyzing the lane width value according to the image information to be analyzed; and then translating the estimated driving trajectory in the plane coordinate system with the lane width value to obtain a safe driving trajectory. The expansion strategy includes obtaining the vehicle model, retrieving the vehicle's width and wheel size values ​​based on the vehicle model, comparing the vehicle's wheel size values ​​with the wheel size values ​​of vehicles providing safe driving trajectories to obtain first ratio information, comparing the vehicle's width value with the width value of vehicles providing safe driving trajectories to obtain second ratio information, combining the first and second ratio information to obtain comprehensive change information, and then calculating the wheel rolling trajectory area by indexing the comprehensive change information based on the safe driving trajectory. It also includes step S4, which has a preset safe distance value. If there is a vehicle in the adjacent lane, the estimated driving trajectory and the safe driving trajectory are extended respectively. It is determined whether the two extended trajectory lines intersect. If they intersect, the intersection point is marked as the collision point in the plane coordinate system. The distance between the vehicle and the adjacent vehicle is calculated based on the collision point, the position of the vehicle and the position of the adjacent vehicle. When the distance between the vehicle and the adjacent vehicle is equal to the safe distance value, the current position of the vehicle is recorded as the undetermined point. The subsequent trajectory is obtained through a self-learning strategy based on the undetermined point. The self-learning strategy includes comparing the size value of the vehicle inserted into the lane with the size value of the initially tracked vehicle based on the image information to be analyzed. If the size value of the vehicle inserted into the lane is less than the size value of the initially tracked vehicle, the historical driving trajectory of the initially tracked vehicle is used as the subsequent trajectory. If the size value of the vehicle inserted into the lane is greater than or equal to the size value of the initially tracked vehicle, the vehicle inserted into the lane is used as the tracking target, and the historical driving trajectory of the tracking target is obtained as the subsequent trajectory.

2. The intelligent vehicle following method according to claim 1, characterized in that: The preprocessing strategy further includes a preprocessing sub-strategy, which includes extracting the starting points of the lane lines on both sides of the lane the vehicle is about to enter as prediction points from the image information to be analyzed; marking the prediction points in a plane coordinate system and indexing the center point of the two prediction points as the ideal point of the vehicle; obtaining the switching point of the center point of the vehicle's front in the plane coordinate system when the vehicle travels straight until it completely leaves the current lane; connecting the origin and the switching point to obtain the straight trajectory; calculating the turning trajectory by using a trajectory planning function formula with the switching point and the ideal point of the vehicle; and connecting the straight trajectory and the turning trajectory to obtain the safe driving trajectory.

3. The intelligent vehicle following method according to claim 2, characterized in that: Step S2 further includes step S21, which involves a preset curvature reference interval. Based on the historical driving trajectory or the estimated driving trajectory, it is determined whether the curvature in the trajectory is greater than the curvature reference interval. If any curvature is greater than the curvature reference interval, the starting point of that trajectory segment is taken as the departure starting point. The departure starting point includes the initial point where the preceding vehicle leaves the current safe driving trajectory. The departure starting point and the ideal point of this vehicle are used to calculate the subsequent trajectory through a trajectory formula.

4. The intelligent vehicle following method according to claim 1, characterized in that: Step S4 further includes step S41, which presets a following distance value. Based on the image information to be analyzed, the distance between the vehicle and the initially tracked vehicle is determined as the actual distance value. If the actual distance value is greater than the following distance value, the vehicle stops at the pending point. If the actual distance value is less than or equal to the following distance value, the vehicle issues a siren and continues to use the historical driving trajectory of the initially tracked vehicle as the subsequent trajectory.

Images