Lane change detection method, electronic device, vehicle and storage medium

Abstract

The present invention discloses a lane change detection method, electronic device, vehicle and storage medium, which are used in the field of vehicle technology. The method includes: obtaining the corner point position of the target vehicle; based on the corner point position of the target vehicle, performing lane change detection on the target vehicle to obtain a lane change detection result. Since the target vehicle usually moves first at the front or rear end when changing lanes, the change of the corner point position is more obvious. The lane change detection based on the corner point position can improve the accuracy of the lane change detection; and the corner point position can reflect the characteristics of the boundary of the target vehicle, that is, the lane change detection is performed under the condition of accurately locating the position of the target vehicle in the lane, so as to further improve the accuracy of the acquired lane change detection result.

Description

Technical Field

[0001] The present invention relates to the field of vehicle technology, and more specifically, to a lane change detection method, an electronic device, a vehicle, and a non-volatile computer-readable storage medium. Background Art

[0002] In the vehicle's autonomous driving technology, the prediction module in the on-board algorithm usually predicts the behavioral intentions (e.g., lane change intentions, etc.) of surrounding road traffic participants (e.g., vehicles, cyclists, and pedestrians) based on the perceived information of the surrounding environment, and obtains the prediction results. The vehicle then makes corresponding control decisions based on the prediction results.

[0003] Therefore, the accuracy of the prediction results has a huge impact on the safety of the vehicle during autonomous driving. Summary of the invention

[0004] In view of this, embodiments of the present application provide a lane change detection method, an electronic device, a vehicle, and a non-volatile computer-readable storage medium, which can improve the accuracy of lane change detection of surrounding vehicles.

[0005] The lane change detection method of the present application includes obtaining a corner point position of a target vehicle; and performing a lane change detection on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result.

[0006] In some embodiments, the corner point position includes a left corner point position and a right corner point position, and the lane change detection is performed on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result, including at least one of the following: based on the left corner point position of the target vehicle, a left lane change detection is performed on the target vehicle to obtain a left lane change detection result, and the left lane change detection result includes whether to change left lanes; based on the right corner point position of the target vehicle, a right lane change detection is performed on the target vehicle to obtain a right lane change detection result, and the right lane change detection result includes whether to change right lanes.

[0007] In some embodiments, the lane change detection is performed on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result, including: when the left lane change detection result is to change left lanes, not performing the right lane change detection; or when the right lane change detection result is to change right lanes, not performing the left lane change detection.

[0008] In certain embodiments, the performing lane change detection on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result includes: when the target vehicle is located on the left side of the current vehicle, based on the right corner point position of the target vehicle, performing right lane change detection on the target vehicle to obtain a right lane change detection result; when the target vehicle is located on the right side of the current vehicle, based on the left corner point position of the target vehicle, performing left lane change detection on the target vehicle to obtain a left lane change detection result.

[0009] In some embodiments, the left corner point is located at the upper left corner of the target vehicle, and the right corner point is located at the upper right corner of the target vehicle.

[0010] In some embodiments, it also includes: determining a first corner point offset ratio of the corner point position relative to a preset reference line of the lane; when the first corner point offset ratio is greater than a first preset ratio threshold, entering the corner point position based on the target vehicle, performing lane change detection on the target vehicle to obtain a lane change detection result.

[0011] In some embodiments, determining a first corner point offset ratio of the corner point position relative to a preset reference line of a lane includes: determining a target distance between the corner point position and the preset reference line and a lane width; and determining the first corner point offset ratio based on the target distance and the lane width.

[0012] In some embodiments, the vehicle includes multiple corner points, and the method further includes: obtaining multiple historical corner point positions of a target corner point among the multiple corner points within a first preset time length; based on the multiple historical corner point positions of the target corner point, determining the corner point offset speed and predicting a second corner point offset ratio of the predicted corner point position after a second preset time length relative to the center line of the lane.

[0013] In certain embodiments, determining the corner point offset speed based on the multiple historical corner point positions of the target corner point and predicting the second corner point offset ratio of the predicted corner point position relative to the center line of the lane after a second preset time period includes: determining the corner point offset speed based on the distances of the multiple historical corner point positions of the target corner point relative to a preset reference line of the lane; fitting the third corner point offset ratios and corresponding acquisition times corresponding to the multiple historical corner point positions of the target corner point to generate a fitting curve; and predicting the second corner point offset ratio based on the fitting curve.

[0014] In some embodiments, the fitting of the third corner point offset ratios and the corresponding acquisition times corresponding to the multiple historical corner point positions of the target corner point to generate a fitting curve includes: calculating the root mean square error of the distance from the fitting point corresponding to the multiple historical corner point positions of each of the target corner points to the fitting curve; and when the root mean square error is greater than a preset error threshold, determining that the multiple historical corner point positions of the target corner point are invalid.

[0015] In some embodiments, the lane change detection is performed on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result, including: based on at least one of the corner point offset speed and the second corner point offset ratio, the lane change detection is performed on the target vehicle to obtain a lane change detection result.

[0016] In some embodiments, the lane change detection result includes whether to change lanes, and the lane change detection on the target vehicle based on at least one of the corner point offset speed and the corner point offset distance includes: when the corner point offset speed is greater than a preset speed threshold, determining that the target vehicle is changing lanes; and / or when the second corner point offset ratio is greater than a second preset ratio threshold, determining that the target vehicle is changing lanes.

[0017] The electronic device of the implementation mode of the present application includes a processor, which is connected to a memory; the memory stores a computer program, and the processor executes the computer program to implement instructions of the lane change detection method described in any of the above implementation modes.

[0018] The vehicle according to the embodiment of the present application includes the electronic device described in any of the above embodiments.

[0019] The non-volatile computer-readable storage medium of the embodiment of the present application includes a computer program. When the computer program is executed by a processor, the processor executes the lane change detection method described in any of the above embodiments.

[0020] The lane change detection method, electronic device, vehicle, and non-volatile computer-readable storage medium of the embodiment of the present application obtain the corner point position of the target vehicle, and then perform lane change detection on the target vehicle based on the corner point position to obtain the lane change detection result. Since the target vehicle usually moves first at the front or rear end when changing lanes, the change in the corner point position is more obvious, that is, the corner point position can reflect the lane change intention of the target vehicle. Therefore, the lane change detection based on the corner point position can improve the accuracy of the lane change detection; and the corner point position can also reflect the characteristics of the boundary of the target vehicle, so that the lane change detection can be performed while accurately locating the position of the target vehicle in the lane, further improving the accuracy of the acquired lane change detection result.

[0021] Additional aspects and advantages of the embodiments of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0023] Figure 1 is a schematic diagram of an application scenario of a lane change detection method in certain embodiments of the present application;

[0024] Figure 2 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0025] Figure 3 is a schematic diagram of a scenario of a lane change detection method according to certain embodiments of the present application;

[0026] Figure 4 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0027] Figure 5 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0028] Figure 6 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0029] Figure 7 is a schematic diagram of a scenario of a lane change detection method according to certain embodiments of the present application;

[0030] Figure 8 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0031] Fig. 9 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0032] Fig.10 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0033] Fig.11 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0034] Fig.12 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0035] Fig.13 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0036] Fig.14 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0037] Fig.15 is a flowchart of a lane change detection method according to certain embodiments of the present application;

[0038] Fig.16 is a schematic diagram of a module of a lane change detection device according to certain embodiments of the present application;

[0039] Fig.17 It is a schematic diagram of the connection status of a non-volatile computer-readable storage medium and a processor in certain embodiments of the present application. DETAILED DESCRIPTION

[0040] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions from beginning to end. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the embodiments of the present application, and cannot be understood as limiting the embodiments of the present application.

[0041] In the field of vehicle autonomous driving, the prediction module in the on-board algorithm is responsible for predicting the future movement trends and behavioral intentions of surrounding road traffic participants (such as vehicles, cyclists and pedestrians) based on surrounding perception information and obtaining prediction results. The quality of the prediction performance plays a vital role in the vehicle's downstream decision-making (i.e., the actual decision made by the vehicle on how to perform driving operations based on the prediction results) and planning modules (such as path planning, etc.). The prediction module is the most important module for interacting with surrounding road traffic participants during autonomous driving.

[0042] The prediction module can be divided into trajectory prediction and behavior prediction according to different prediction tasks. Among them, behavior prediction generally refers to the prediction of lane change behavior of surrounding traffic participants. During the driving process of the autonomous driving vehicle, the downstream decision module will execute different driving strategies based on the behavior prediction results of the surrounding traffic participants. For example, when the behavior prediction results of the surrounding traffic participants are obtained, the downstream decision module can make a driving strategy to control the vehicle to perform defensive deceleration.

[0043] It is understandable that the accuracy of behavior prediction will greatly affect the safety of autonomous driving.

[0044] In order to solve the above technical problems, an embodiment of the present application provides a lane change detection method.

[0045] The lane change detection method of the present application will be described in detail below:

[0046] See also Figure 1 The lane change detection method provided in this application can be applied to Figure 1 In the application scenario shown, the lane change detection method can be applied to a vehicle 100 , which is any vehicle 100 that can predict the lane change of surrounding vehicles (such as the target vehicle 200 ).

[0047] See also Figure 1 and Figure 2 , an embodiment of the present application provides a lane change detection method, the lane change detection method comprising:

[0048] Step 011: Get the corner point position of the target vehicle.

[0049] The target vehicle may be other vehicles located around the vehicle while the vehicle is driving.

[0050] The corner point may be at least one of the four corner points after the vehicle body is projected onto the ground. The corner point position may be the coordinates of the corner point. For example, see Figure 3 The corner point position may be the corner point position of any one of the four corner points (corner point A, corner point B, corner point C, and corner point D) obtained after the target vehicle 200 is projected on the ground.

[0051] The vehicle can collect information about traffic participants (such as pedestrians, other vehicles, etc.) around the vehicle through sensors (such as infrared sensors, visual sensors, radar sensors, etc.), determine the target vehicle among the surrounding traffic participants, and then determine the corner point position of the target vehicle based on the corner points.

[0052] See also Figure 4 Optionally, the lane change detection method further includes:

[0053] Step 013: Obtain the position coordinates of the corner point in the vehicle coordinate system;

[0054] Step 014: Based on the mapping relationship between the vehicle coordinate system and the centerline coordinate system of the lane, each position coordinate is converted into a corresponding corner point position. The centerline coordinate system is established based on the centerline of the lane where the target vehicle is located.

[0055] Specifically, the position coordinates (veh corner x, veh corner y) of the corner point of the target vehicle in the vehicle coordinate system (VCS coordinate system) can be obtained through the vehicle's sensors; or, the position coordinates of the corner point of the target vehicle can be calculated based on the relevant information of the target vehicle by obtaining relevant information of the target vehicle, including length information (obs length), width information (obs width), center coordinate information of the target vehicle (veh center x, veh center y) and yaw angle information of the target vehicle (obs yaw). For example, the position coordinates (veh corner x1, veh corner y1) of the left front corner point of the target vehicle are calculated as an example, where vehcorner x1 can be obtained by the following calculation:

[0056]

[0057] The veh corner y1 can be calculated as follows:

[0058]

[0059] Based on the center line of the lane, the center line coordinate system of the lane is established, that is, the Frenet coordinate system (the Frenet coordinate system is established based on the center line of the lane where the target vehicle is located, that is, the center line of the lane is used as the reference line, and the coordinate system is established based on the tangent vector and normal vector of the reference line. For example, see Figure 1 , a Frenet coordinate system with abscissa S and ordinate L is established based on the center line of the lane, where the left side of the center line of the lane is positive and the right side is negative). Through the mapping relationship between the vehicle's vcs coordinate system and the Frenet coordinate system established based on the center line of the lane, the position coordinates of the corner point in the vehicle's vcs coordinate system are converted to the corner point position in the Frenet coordinate system established based on the center line of the lane.

[0060] Step 012: Based on the corner point position of the target vehicle, a lane change detection is performed on the target vehicle to obtain a lane change detection result.

[0061] Among them, the lane change detection result may include whether the target vehicle changes lanes, and the vehicle can make corresponding control strategies based on the lane change detection result of the target vehicle.

[0062] Specifically, by analyzing the change in the position of the corner point of the target vehicle (for example, by obtaining the position of the corner point within a preset time length, analyzing the moving direction and moving speed of the corner point of the target vehicle), a lane change detection is performed on the target vehicle (for example, analyzing the movement trend of the corner point to predict the position of the corner point, etc.), so as to obtain a lane change detection result (for example, whether the target vehicle changes lanes). For example, in the case where the corner point of the target vehicle moves toward the lane line position at a faster movement speed, according to the movement trend of the corner point, it is predicted that the corner point may reach the lane line position after the preset time length, and therefore, the lane change detection result of the target vehicle changing lanes can be obtained to indicate that the target vehicle changes lanes.

[0063] Among the various lane change prediction algorithms currently available, whether they are rule-based or deep learning-based, their input data often only focuses on the motion characteristics of the vehicle center (the center position of the vehicle body). For example, based on the historical position changes of the vehicle center point, the position changes of the vehicle center point relative to the lane line, etc. In real road driving scenarios, due to the differences in vehicle types, sizes, driver habits, and traffic conditions, algorithms that only rely on the vehicle center coordinates for lane change prediction cannot achieve good prediction results for all driving scenarios.

[0064] For example, take the algorithm for lane change prediction based on the movement trend of the center point of the vehicle as an example. Assuming that the direction of vehicle movement is vertical, the width of a large vehicle with a larger body size and a small vehicle with a smaller body size are different, and the dividing point of lane change (or lateral movement in the lane) is different. For example, when the center point of the vehicle is obtained to have moved 1 meter (m) laterally, for the small car, the small car is still in the lane, and the driver may just want to correct the centering of the driving and has no intention of changing lanes; while for the large car, the body of the large car has crossed the line and is in the process of changing lanes. Therefore, in order to improve the accuracy of the vehicle's lane change prediction, the current lane change prediction algorithm also needs to set different lane change thresholds for vehicles of various body sizes and various driving scenarios (lanes of various widths). On the one hand, it increases the difficulty of algorithm development, and on the other hand, when predicting lane changes, the algorithm is very likely to fall into the game of missed reports and false reports, and it is difficult to achieve high precision and high recall at the same time.

[0065] Lane change detection based on the corner point position of the target vehicle can not only reduce the difficulty of algorithm development, but also improve the accuracy of detection. This is because the corner point can reflect the characteristics of the boundary of the target vehicle. Lane change detection based on the corner point position can accurately locate the position and movement trend of the target vehicle in the lane. When the target vehicle changes lanes, the movement of the front or rear of the target vehicle can be captured more accurately, thereby improving the accuracy of lane change detection.

[0066] In this way, by obtaining the corner point position of the target vehicle, and then based on the corner point position of the target vehicle, lane change detection is performed on the target vehicle to obtain the lane change detection result. Since the change of the corner point position is more obvious when the target vehicle changes lanes, it can better reflect the lane change intention of the target vehicle. Therefore, lane change detection based on the corner point position can improve the accuracy of lane change detection; and, based on the corner point position, the movement of the front or rear of the target vehicle can be accurately captured, and the corner point position can reflect the characteristics of the boundary of the target vehicle, that is, lane change detection can be performed while accurately locating the position of the target vehicle in the lane, further improving the accuracy of the acquired lane change detection result.

[0067] Compared with the method of lane change detection based on the center position of the vehicle, lane change detection based on the corner point position can detect target vehicles of different body sizes, that is, lane change detection of vehicles in complex scenarios is more robust.

[0068] In some embodiments, the corner point position includes a left corner point position and a right corner point position. Step 012: based on the corner point position of the target vehicle, performing lane change detection on the target vehicle to obtain a lane change detection result includes at least one of the following:

[0069] (1) Based on the left corner point position of the target vehicle, a left lane change detection is performed on the target vehicle to obtain a left lane change detection result, where the left lane change detection result includes whether a left lane change is performed;

[0070] (2) Based on the right corner point position of the target vehicle, a right lane change detection is performed on the target vehicle to obtain a right lane change detection result. The right lane change detection result includes whether to perform a right lane change.

[0071] Please refer again to Figure 3 The corner points of the target vehicle may include a left corner point A and a left corner point C, a right corner point B and a right corner point D, wherein the positions of the left corner point A and the left corner point C correspond to the left corner point position, and the positions of the right corner point B and the right corner point D correspond to the right corner point position.

[0072] Optionally, the left corner point is located at the upper left corner of the target vehicle, and the right corner point is located at the upper right corner of the target vehicle.

[0073] Please refer again to Figure 3 , the left corner point is located at the upper left corner of the target vehicle (i.e., left corner point A), and the right corner point is located at the upper right corner of the target vehicle (i.e., right corner point B). When a vehicle changes lanes by turning, the front wheels of the vehicle are generally used to control the turning, and the rear wheels of the vehicle follow the turning. Therefore, the position of the target vehicle can be more accurately obtained through the corner point position of the left corner point A at the upper left corner of the target vehicle and the corner point position of the right corner point B at the upper right corner of the target vehicle, thereby improving the accuracy of lane change detection.

[0074] Among them, the left lane change can be that the target vehicle changes lanes from right to left, from the right lane to the left lane, and the right lane change can be that the target vehicle changes lanes from left to right, from the left lane to the right lane.

[0075] Specifically, when the target vehicle changes lanes, the target vehicle's driving direction is usually adjusted by turning the front or rear of the vehicle. For example, when the target vehicle changes lanes to the left, the front of the target vehicle may turn left, resulting in a significant change in the position of the left corner point; when changing lanes to the right, the front of the target vehicle may turn right, resulting in a significant change in the position of the right corner point. Therefore, the target vehicle can be detected for left lane change based on the position of the left corner point of the target vehicle, and the target vehicle can be detected for right lane change based on the position of the right corner point of the target vehicle to obtain the corresponding left lane change detection result or right lane change detection result, thereby determining whether the vehicle is changing lanes to the left or to the right, and improving the accuracy of the prediction of the target vehicle's lane change intention.

[0076] It can be understood that when the target vehicle changes lanes, each corner point moves with the target vehicle. For example, when the target vehicle changes lanes to the left, the right corner point of the target vehicle will also move to the left. Therefore, the target vehicle can also be detected for right lane change based on the position of the left corner point of the target vehicle, and the target vehicle can be detected for left corner point based on the position of the right corner point of the target vehicle, and the lane change intention of the target vehicle can be judged based on the corresponding left lane change detection result or right lane change detection result.

[0077] See also Figure 5 In some embodiments, step 012: performing lane change detection on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result includes:

[0078] Step 0121: when the left lane change detection result indicates that a left lane change is to be performed, no right lane change detection is performed; or

[0079] Step 0122: When the right lane change detection result indicates that a right lane change is to be performed, a left lane change detection is not performed.

[0080] Specifically, left lane change detection and right lane change detection are performed on the target vehicle. When the detection result of the first lane change direction is a lane change, the lane change detection in the other direction is not performed. Take the example of first performing left lane change detection on the target vehicle and then performing right lane change detection. When performing lane change detection on the target vehicle, if the left lane change detection result of the target vehicle is a left lane change, it is determined that the target vehicle is changing lanes. In order to save computing power, it is not necessary to perform right lane change detection on the target vehicle. If the target vehicle does not change lanes to the left, the target vehicle is then subjected to right lane change detection to confirm whether the target vehicle is performing right lane change detection, thereby improving the accuracy of lane change prediction.

[0081] See also Figure 6 In some embodiments, step 012: performing lane change detection on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result includes:

[0082] Step 0123: When the target vehicle is located on the left side of the current vehicle, based on the right corner point position of the target vehicle, a right lane change detection is performed on the target vehicle to obtain a right lane change detection result;

[0083] Step 0124: When the target vehicle is located on the right side of the current vehicle, a left lane change detection is performed on the target vehicle based on the left corner point position of the target vehicle to obtain a left lane change detection result.

[0084] Specifically, see Figure 7 , when the target vehicle 200 is located on the left side of the vehicle 100, if the target vehicle 200 changes lanes to the right, it will enter the lane where the vehicle 100 is located, which has a greater impact on the self-driving decision of the vehicle 100. Therefore, when the target vehicle 200 is located on the left side of the vehicle 100, the right corner point position of the target vehicle (such as Figure 7 The right corner point position of the target vehicle 200 is used to detect the right lane change of the target vehicle 200, so as to obtain the right lane change detection result of whether the target vehicle 200 changes lanes to the right and enters the lane where the vehicle 100 is located. Similarly, when the target vehicle 200 is located on the right side of the vehicle 100, if the target vehicle 200 changes lanes to the left at this time, it will enter the lane where the vehicle 100 is located, which has a greater impact on the self-driving decision of the vehicle 100. When the target vehicle 200 is located on the right side of the vehicle 100, the left corner point position of the target vehicle 200 (such as Figure 7 The left corner point position of the left corner point A in the image is determined by the detection result of the left lane change of the target vehicle to obtain the right lane change detection result of whether the target vehicle changes lanes left and enters the lane where the vehicle is located. In this way, not only the accuracy of the prediction of the lane change detection result can be improved, but also computing power can be saved.

[0085] See also Figure 8 In some embodiments, the lane change detection method further includes:

[0086] Step 015: Determine the first corner point offset ratio of the corner point position relative to the preset reference line of the lane; when the first corner point offset ratio is greater than the first preset ratio threshold, proceed to step 012: Based on the corner point position of the target vehicle, perform lane change detection on the target vehicle to obtain a lane change detection result.

[0087] Among them, the preset reference line can be the center line of the lane, the lane line of the lane, etc. This application takes the preset reference line as the center line of the lane as an example for explanation.

[0088] Among them, the first corner point offset ratio can be used to characterize the corner point position of the target vehicle, and the degree of deviation from a preset reference line of the lane where the corner point is located (for example, the center line of the lane). The larger the first corner point offset ratio is, the more the corner point of the target vehicle deviates from the center line of the lane (that is, the closer it is to the lane line).

[0089] The first preset ratio threshold may be an empirical value, a historical value, etc. For example, taking the leftward offset as positive and the rightward offset as negative as an example, the first preset ratio threshold may be 0.2 and -0.2. When the first corner point offset ratio of the corner point position of the target vehicle relative to the center line of the lane is greater than the first preset ratio threshold, it may be considered that the target vehicle may have an intention to change lanes.

[0090] Specifically, the first corner point offset ratio of the corner point position relative to the center line of the lane where the corner point is located can be calculated. When the first corner point offset ratio is greater than a first preset ratio threshold, the target vehicle is away from the center line of the lane and may have an intention to change lanes. At this time, lane change detection can be performed based on the corner point position of the target vehicle.

[0091] It is understandable that if the target vehicle is predicted to change lanes based only on the moving distance of the corner point position, the moving distance setting will be difficult to control due to the increased complexity of the scene (for example, the scene of lane change detection for vehicles of different sizes or lanes of different widths). If the moving distance of the corner point is relied upon alone, the accuracy of lane change prediction may be reduced. Therefore, it is necessary to further distinguish the corner point position from vehicles of different types and sizes, and further distinction will increase the difficulty of algorithm development and calculation.

[0092] Therefore, by calculating the degree of deviation of the corner point position relative to the center line of the lane (i.e., the first corner point offset ratio), whether to perform lane change detection is determined according to the degree of deviation between the corner point and the center line of the lane (the degree of proximity between the corner point and the lane line). In other words, the moving distance of the target vehicle is standardized and converted into the degree of deviation of the target vehicle relative to the center line of the lane. This can reduce the complexity of the vehicle lane change prediction algorithm. For target vehicles of different sizes and lanes of different widths, the first corner point offset ratio can be used to make corresponding predictions, which improves the versatility of the method. In addition, the first corner point offset ratio can better characterize the lane change intention of the target vehicle, and can improve the accuracy of the lane change detection results.

[0093] It should be noted that see Figure 1, when the target vehicle changes lanes from lane 1 to lane 2, when the corner point position is calculated based on the lane centerline of lane 2, the lane change detection result may be that the target vehicle is still going to change lanes, but at this time the target vehicle has no intention of continuing to change lanes. Therefore, by setting a first preset proportional threshold (such as 0.2 and -0.2), when the corner point crosses the centerline of the lane after entering the new lane and deviates from the lane centerline by 0.2 or -0.2 of the lane width, the lane change detection is triggered again to avoid false alarms of the end of the lane change and the continuous lane change of the target vehicle.

[0094] See also Fig. 9 Optionally, step 015: determining a first corner point offset ratio of the corner point position relative to a preset reference line of the lane includes:

[0095] Step 0151: Determine the target distance between the corner point position and the preset reference line and the lane width;

[0096] Step 0152: Based on the target distance and lane width, determine the offset ratio of the first corner point.

[0097] Specifically, see Figure 1 , the left lane line is the lane line (left lane line) L1 on the left side of the lane where the corner point of the target vehicle 200 is located, and the right lane line is the lane line (right lane line) R1 on the right side. The position coordinates of the corner point of the target vehicle 200 can be obtained through the sensor of the vehicle 100, and then the position coordinates of the corner point are converted to the corner point position of the Frenet coordinate system. Then, the target distance corner l (such as Figure 1 d1 as shown), and then obtain the second distance (left lane line) between the corner point position and the left lane line L1 (as shown Figure 1 d2 shown), and the third distance between the corner point position and the right lane line R1 (right lane line) (as shown Figure 1 As shown in FIG. 3 , the second distance d2 and the third distance d3 are added to determine the lane width W; and then the ratio of the target distance d1 and the lane width W is calculated to determine the first corner point offset ratio corresponding to the corner point.

[0098] For example, the first distance corresponding to the left corner point of the target vehicle is recorded as left corner center l, and the first distance corresponding to the right corner point is recorded as right corner center l. The first corner point offset ratio of the left corner point, that is, the left corner point offset ratio (left corner lateral ratio), can be calculated as follows:

[0099] ;

[0100] For example, see Figure 1 , , where W=d2+d3;

[0101] The first corner offset ratio of the right corner point, that is, the right corner offset ratio (right corner lateralratio), can be calculated as follows:

[0102]

[0103] Since the Frenet coordinate system established based on the center line of the lane takes the center line of the lane as the origin, the left direction is defined as positive, and the right direction is defined as negative, the distance between the corner point position and the center line of the lane is positive or negative, and the first corner point offset ratio is also positive or negative. When the first corner point offset ratio of any corner point is 0, the corner point is located on the center line of the lane; when the first corner point offset ratio of any corner point is 0.5, the corner point is located on the left lane line of the lane; when the first corner point offset ratio of any corner point is -0.5, the corner point is located on the right lane line of the lane.

[0104] The distance between the corner point and the lane line is converted into the first corner point offset ratio, and the positive and negative signs are used to distinguish whether the corner point is close to the left lane line or the right lane line of the lane. The first corner point offset ratio can eliminate the influence of different lanes (for example, different widths) and different target vehicles (for example, different vehicle widths) on the value of the offset distance. The offset distance is normalized to the degree of deviation of the target vehicle relative to the center line of the lane. The lane change prediction is performed based on the degree of deviation, which can better characterize the lane change intention of the target vehicle and improve the accuracy of the lane change prediction.

[0105] It can be understood that the second distance between the corner point position and the left lane line and the third distance between the corner point position and the right lane line can be calculated by respectively obtaining the mapping relationship between the vcs coordinates of the vehicle and the Frenet coordinate system established based on the left lane line, and the mapping relationship between the vcs coordinates of the vehicle and the Frenet coordinate system established based on the right lane line, and then adding the second distance and the third distance to determine the lane width, wherein when calculating the second distance and the third distance, the units need to be unified (for example, both are meters (m)).

[0106] See also Fig.10 In some embodiments, the vehicle includes a plurality of corner points, and the method further includes:

[0107] Step 016: Acquire multiple historical corner point positions of a target corner point among multiple corner points within a first preset time length;

[0108] Step 017: Based on multiple historical corner point positions of the target corner point, determine the corner point offset speed and predict a second corner point offset ratio of the predicted corner point position relative to the center line of the lane after a second preset time period.

[0109] The first preset duration and the second preset duration may be set based on experience values, historical values, etc. For example, both the first preset duration and the second preset duration may be 1 second (s), 2s, 3s, etc. For example, the first preset duration may be 1500ms, 2000ms, 2500ms, etc. The first preset duration and the second preset duration may be the same or different.

[0110] It generally takes about 500ms of link delay for the vehicle to respond to the lane change prediction result. Therefore, the second preset time length may be greater than 500ms, for example, 600ms, 700ms, etc., so that the vehicle can make corresponding control strategies as soon as possible based on the lane change prediction result.

[0111] Specifically, the corner point offset speed can be determined by obtaining multiple historical corner point positions of a target corner point (such as the left front corner point among the four corner points of the target vehicle) within a first preset time period, and according to the multiple historical corner point positions, for example, please refer again to Figure 1 , the corner point offset speed of the target corner point can be determined by obtaining the distance difference between the historical corner point position a1 and the historical corner point position a2 of the corner point A (for example, the distance difference between the target distances of the two historical corner point positions) and the time difference between the timestamps of the two historical corner point positions, and by the ratio of the distance difference to the time difference. Then, according to the corner point offset speed and the corner point offset direction of the target corner point, the predicted corner point position of the target corner point after the second preset time length and the second corner point offset ratio corresponding to the predicted corner point position are predicted.

[0112] To improve data stability, see Fig.11 In some embodiments, the lane change detection method further includes:

[0113] Step 018: Filtering the data of multiple historical corner point positions;

[0114] Step 012: Based on the corner point position of the target vehicle, lane change detection is performed on the target vehicle to obtain a lane change detection result, including:

[0115] Step 0121: Based on the filtered historical corner point positions, lane change detection is performed on the target vehicle to obtain a lane change detection result.

[0116] Among them, the filtering can be mean filtering, extreme value filtering and smoothing filtering, etc. This application takes mean filtering as an example.

[0117] Specifically, due to the inherent performance upper limit of the vehicle's perception equipment (for example, sensors (such as lidar, cameras, etc.)), the limitations of the detection algorithm and multiple sensor data fusion processing algorithms, the perception of the target vehicle may be unstable. For example, the perceived numerical results contain jumps (discontinuous counting) and noise data. Performing lane change prediction when the perception is unstable may cause the accuracy of the lane change detection results to be affected due to abnormal values ​​in certain frames. Therefore, it is possible to filter the data of multiple historical corner point positions (such as mean filtering), and then perform lane change detection based on the filtered corner point positions. Compared with the current direct use of perception data for lane change detection, the present application is based on filtered data. It can minimize the impact of data fluctuations and data jumps on the lane change detection results of the vehicle's perception of the target vehicle without affecting the accuracy of the lane change detection, thereby improving the accuracy of the lane change detection.

[0118] For example, taking the calculation of the corresponding first corner point offset ratio according to multiple historical corner point positions and the mean filtering based on the first corner point offset ratio as an example, first define the time window length of the mean filter, because the longer the time window, the better the filtering smoothing effect, but the greater the phase delay; the shorter the time window, the smaller the phase delay, but the worse the filtering smoothing effect. Therefore, the time window length needs to be determined by comprehensively considering the length of the data of the first corner point offset ratio. Taking the data of the first corner point offset ratio with a duration of 2000ms as an example, according to multiple comparative experiments, the time window length of the mean filter can be determined as 200ms, and then the abnormal values ​​(such as the maximum and minimum values) of the first corner point offset ratio within the 200ms time window are removed; finally, the first corner point offset ratio within 2000ms is mean filtered based on the filtering algorithm to obtain a smoother and more stable first corner point offset ratio, thereby improving the accuracy of lane change detection.

[0119] See also Fig.12In some embodiments, step 017: determining the corner point offset speed based on the multiple corner point positions of the target corner point and predicting a second corner point offset ratio of the predicted corner point position relative to the center line of the lane after a second preset time period includes:

[0120] Step 0171: determining a corner point offset speed based on distances of multiple historical corner point positions of the target corner point relative to a preset reference line of the lane;

[0121] Step 0172: fitting the third corner point offset ratios and corresponding acquisition times corresponding to multiple historical corner point positions of the target corner point to generate a fitting curve;

[0122] Step 0173: Based on the fitting curve, predict the offset ratio of the second corner point.

[0123] Specifically, based on the filtered data of multiple historical corner point positions, by obtaining the distances of multiple historical corner point positions of the target corner point relative to the center line of the lane where the corner point is located, the distance difference between any two historical corner point positions can be calculated (or the ratio of the distance difference to the lane width can be calculated), and the displacement distance (or displacement ratio) can be determined, and the corner point offset speed can be determined according to the displacement distance (or displacement ratio) and the time difference of collecting the two historical corner point positions. By fitting the third corner point offset ratio corresponding to the multiple historical corner point positions and the acquisition time corresponding to each historical corner point position, the acquisition time (the number of frames collected) is used as the independent variable (x), and the third corner point offset ratio is used as the dependent variable (y), and the least squares fitting is performed based on the independent variable x and the dependent variable y to obtain a fitting curve (y=kx+b), wherein the slope k of the fitting curve is the movement speed of the first corner point offset ratio, and the intercept b can be the first corner point offset ratio corresponding to the starting corner point position of the target corner point. Therefore, based on the fitting curve, the second preset time length is substituted as x into y=kx+b, and the second corner point offset ratio after the second preset time length can be predicted.

[0124] The lane changing time of a vehicle when changing lanes (starting from the moment the driver turns the steering wheel and ending from the moment the corner of the vehicle presses against the lane line of the lane in which it is located, and the time from the start to the end) is generally 400-1000 milliseconds (ms). Taking into account the fitting effect of the fitting curve, the present application can set the first preset time to any time within the time range of 400ms-1000ms, for example, 500ms, 600ms, 700ms, etc., to avoid the problem of poor fitting effect due to a fitting time window that is too long and unclear motion patterns due to a fitting time window that is too short.

[0125] See also Fig.13 In some implementations, step 0172: fitting the third corner point offset ratios and the corresponding acquisition times corresponding to the multiple historical corner point positions of the target corner point to generate a fitting curve includes:

[0126] Step 01721: Calculate the root mean square error of the distance from the fitting point to the fitting curve corresponding to the multiple historical corner point positions of each target corner point;

[0127] Step 01722: When the root mean square error is greater than a preset error threshold, it is determined that the positions of multiple historical corner points of the target corner point are invalid.

[0128] The preset error threshold may be a historical value or an empirical value.

[0129] Specifically, after the target vehicle starts to change lanes, the corner point position of the target vehicle generally shows a trend of changing at a constant speed or accelerating close to the lane line. When the change of the corner point position fluctuates, it can be considered that the target vehicle is in a driving scene where it shakes violently from side to side or is difficult to perceive (perception is unstable). This type of driving scene usually does not belong to the scene where the target vehicle changes lanes. Therefore, in order to improve the accuracy of detection and avoid false positives, the root mean square error of the distance from the third corner point offset ratio corresponding to the multiple historical corner point positions of each target corner point to the fitting curve can be calculated. The root mean square error can be used to characterize the degree of approximation between the fitting result of the fitting curve and the actual corner point position. Then, by setting a preset error threshold, when the root mean square error is greater than the preset error threshold, it can be considered that the data fluctuations of the multiple historical corner point positions of the target corner point are large, which does not conform to the linear law when the target vehicle changes lanes. The lane change detection result obtained by detecting based on the corner point position with large data fluctuations may cause the lane change detection result to be presented as an erroneous result. Therefore, lane change detection should not be performed based on these data. Instead, multiple historical corner point positions of the target corner point are determined to be invalid, so as to filter out data that may cause errors in the lane change detection result and improve the accuracy of the lane change detection.

[0130] See also Fig.14 In some embodiments, step 012: performing lane change detection on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result includes:

[0131] Step 0125: Based on at least one of the corner point offset speed and the second corner point offset ratio, perform lane change detection on the target vehicle to obtain a lane change detection result.

[0132] Specifically, the corner point offset speed can characterize the movement state of the target vehicle in the lateral direction of the lane line, and the second corner point offset ratio can characterize the degree to which the target vehicle is away from the lane center line and close to the lane line. When the target vehicle changes lanes, the target vehicle usually approaches the lane line at a certain lateral movement speed. Therefore, lane change detection can be performed on the target vehicle based on at least one of the corner point offset speed and the second corner point offset ratio to obtain a lane change detection result.

[0133] See also Fig.15 Optionally, the lane change detection result includes whether to change lanes. Step 0125: Based on at least one of the corner point offset speed and the corner point offset distance, lane change detection is performed on the target vehicle, including:

[0134] Step 01251: When the corner point offset speed is greater than a preset speed threshold, determine that the target vehicle is changing lanes; and / or when the second corner point offset ratio is greater than a second preset ratio threshold, determine that the target vehicle is changing lanes.

[0135] The lane change detection result includes whether to change lanes. For example, the lane change detection result includes whether the target vehicle changes lanes and whether the target vehicle does not change lanes.

[0136] The preset speed threshold and the second preset ratio threshold may be historical values ​​or empirical values, etc. The second preset ratio threshold may be 0.5, that is, the corner point position is located on the lane line.

[0137] Specifically, for some target vehicles with larger bodies or scenarios where the driver has a habit of driving close to the line, the target vehicle may be in a state of driving close to the line (lane line) for a long time. In this state, the target vehicle has a certain corner point offset speed, but this corner point offset speed cannot indicate that the target vehicle wants to change lanes. When the target vehicle has the intention to change lanes, the corner point offset speed is usually greater than the corner point offset speed in the state of driving close to the line. Therefore, by setting a preset speed threshold, when the corner point offset speed is greater than the preset speed threshold, that is, when the corner point is in a high lateral motion state, the motion is considered to be the lane change motion of the target vehicle, and the target vehicle is determined to change lanes, so as to avoid misjudgment of lane change prediction caused by low-speed lateral movement in the state of driving close to the line.

[0138] When the target vehicle moves sideways or shakes laterally in the lane (for example, the driver wants to adjust the centering of the target vehicle), the second corner point offset ratio of the target corner point after the second preset time period is usually less than 0.5, that is, the target vehicle usually does not touch the lane line. Therefore, when the second corner point offset ratio of the target corner point after the second preset time period is predicted to be greater than the second preset ratio threshold (such as 0.5), that is, after the second preset time period, the target corner point is located on the lane line, and then the target vehicle is determined to change lanes, it can avoid the lane change false alarm caused by the target vehicle moving sideways or shaking laterally in the lane, and improve the accuracy of the lane change prediction result.

[0139] See also Fig.16In order to better implement the lane change detection method of the embodiment of the present application, the embodiment of the present application also provides a lane change detection device 300. The lane change detection device 300 may include an acquisition module 301 and a detection module 302. The acquisition module 301 is used to acquire the corner point position of the target vehicle; the detection module 302 is used to perform lane change detection on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result.

[0140] In one embodiment, the corner point position includes a left corner point position and a right corner point position, and the detection module 302 is specifically used for at least one of the following: based on the left corner point position of the target vehicle, performing a left lane change detection on the target vehicle to obtain a left lane change detection result, and the left lane change detection result includes whether to change left lanes; based on the right corner point position of the target vehicle, performing a right lane change detection on the target vehicle to obtain a right lane change detection result, and the right lane change detection result includes whether to change right lanes.

[0141] In one embodiment, the detection module 302 is further configured to not perform right lane change detection when the left lane change detection result indicates that a left lane change is to be performed; or not perform left lane change detection when the right lane change detection result indicates that a right lane change is to be performed.

[0142] In one embodiment, the detection module 302 is also specifically used to perform right lane change detection on the target vehicle based on the right corner point position of the target vehicle when the target vehicle is located on the left side of the current vehicle to obtain a right lane change detection result; when the target vehicle is located on the right side of the current vehicle, perform left lane change detection on the target vehicle based on the left corner point position of the target vehicle to obtain a left lane change detection result.

[0143] In one embodiment, the lane change detection device 300 also includes a first determination module 303, which is used to determine a first corner point offset ratio of the corner point position relative to a preset reference line of the lane; when the first corner point offset ratio is greater than a first preset ratio threshold, enter into a lane change detection based on the corner point position of the target vehicle to obtain a lane change detection result.

[0144] In one embodiment, the first determination module 303 is further configured to determine a target distance between the corner point position and a preset reference line and a lane width; and determine a first corner point offset ratio based on the target distance and the lane width.

[0145] In one embodiment, the vehicle includes multiple corner points, and the lane change detection device 300 also includes a second determination module 304, which is used to obtain multiple historical corner point positions of a target corner point among the multiple corner points within a first preset time length; based on the multiple historical corner point positions of the target corner point, determine the corner point offset speed and predict the second corner point offset ratio of the predicted corner point position after the second preset time length relative to the center line of the lane.

[0146] In one embodiment, the second determination module 304 is specifically used to determine the corner point offset speed based on the distances of multiple historical corner point positions of the target corner point relative to a preset reference line of the lane; fit the third corner point offset ratios and corresponding acquisition times corresponding to the multiple historical corner point positions of the target corner point to generate a fitting curve; and predict the second corner point offset ratio based on the fitting curve.

[0147] In one embodiment, the second determination module 304 is specifically used to calculate the root mean square error of the distance from the fitting point to the fitting curve corresponding to multiple historical corner point positions of each target corner point; when the root mean square error is greater than a preset error threshold, it is determined that the multiple historical corner point positions of the target corner point are invalid.

[0148] In one embodiment, the detection module 302 is further configured to perform lane change detection on the target vehicle based on at least one of the corner point offset speed and the second corner point offset ratio to obtain a lane change detection result.

[0149] In one embodiment, the lane change detection result includes whether to change lanes. The detection module 302 is specifically used to determine whether the target vehicle changes lanes when the corner point offset speed is greater than a preset speed threshold; and / or determine whether the target vehicle changes lanes when the second corner point offset ratio is greater than a second preset ratio threshold.

[0150] The lane change detection device 300 is described above from the perspective of functional modules in conjunction with the accompanying drawings. The functional modules can be implemented in hardware form, can be implemented in software form, or can be implemented in combination with hardware and software modules. Specifically, the steps of the method embodiments in the embodiments of the present application can be completed by hardware integrated logic circuits in the processor and / or software form instructions. The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed by a hardware encoding processor, or can be executed by a combination of hardware and software modules in the encoding processor. Optionally, the software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps in the above method embodiments in combination with its hardware.

[0151] Please refer again Figure 1 The vehicle 100 may include a processor 10 and a memory 20, the processor 10 is connected to the memory 20, the memory 20 stores a computer program 610, and the processor 10 executes the computer program 610 to implement any of the lane change detection methods described above. For the sake of brevity, it will not be repeated here.

[0152] The electronic device of the implementation mode of the present application may include a processor, the processor is connected to a memory, a computer program is stored in the memory, and the processor executes the computer program to implement any one of the lane change detection methods described above, which will not be described here for the sake of brevity.

[0153] Among them, the electronic device can be used as a processor or controller of the vehicle (for example, the electronic device is used as a central controller of the vehicle), and the electronic device can be installed in the vehicle so that the vehicle can implement the lane change detection method of any of the above-mentioned embodiments through the electronic device.

[0154] See also Fig.17 The embodiment of the present application further provides a computer-readable storage medium 600 on which a computer program 610 is stored. When the computer program 610 is executed by a processor 620, the steps of the lane change detection method of any of the above-mentioned embodiments are implemented. For the sake of brevity, they are not repeated here.

[0155] In the description of this specification, the descriptions with reference to the terms "certain embodiments", "in an example", "exemplarily", etc., mean that the specific features, structures, materials or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0156] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0157] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A lane change detection method, characterized in that: include: Acquire a corner point position of a target vehicle, wherein the vehicle includes a plurality of corner points; Determine the target distance between the corner point position and a preset reference line of the lane and the lane width; Based on the target distance and the lane width, determining a first corner point offset ratio of the corner point position relative to a preset reference line of the lane; Acquire multiple historical corner point positions of a target corner point among the multiple corner points within a first preset time length; Determining a corner point offset speed based on distances of a plurality of historical corner point positions of the target corner point relative to a preset reference line of the lane; Fitting the third corner point offset ratios and the corresponding acquisition times corresponding to the plurality of historical corner point positions of the target corner point to generate a fitting curve; Based on the fitting curve, predicting the offset ratio of the second corner point; When the first corner point offset ratio is greater than a first preset ratio threshold, performing lane change detection on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result; Wherein, performing lane change detection on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result includes: Based on at least one of the corner point offset speed and the second corner point offset ratio, lane change detection is performed on the target vehicle to obtain a lane change detection result.

2. The lane change detection method according to claim 1, characterized in that: The corner point positions include a left corner point position and a right corner point position, and the lane change detection is performed on the target vehicle based on the corner point positions of the target vehicle to obtain a lane change detection result, including at least one of the following: Based on the left corner point position of the target vehicle, performing a left lane change detection on the target vehicle to obtain a left lane change detection result, wherein the left lane change detection result includes whether to perform a left lane change; Based on the right corner point position of the target vehicle, a right lane change detection is performed on the target vehicle to obtain a right lane change detection result, where the right lane change detection result includes whether to perform a right lane change.

3. The lane change detection method according to claim 2, characterized in that: The performing lane change detection on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result includes: When the left lane change detection result indicates that a left lane change is to be performed, not performing the right lane change detection; or When the right lane change detection result indicates that a right lane change is to be performed, the left lane change detection is not performed.

4. The lane change detection method according to claim 2, characterized in that: The performing lane change detection on the target vehicle based on the corner point position of the target vehicle to obtain a lane change detection result includes: In a case where the target vehicle is located on the left side of the current vehicle, based on the right corner point position of the target vehicle, performing right lane change detection on the target vehicle to obtain a right lane change detection result; In a case where the target vehicle is located on the right side of the current vehicle, a left lane change detection is performed on the target vehicle based on the left corner point position of the target vehicle to obtain a left lane change detection result.

5. The lane change detection method according to any one of claims 2 to 4, characterized in that: The left corner point is located at the upper left corner of the target vehicle, and the right corner point is located at the upper right corner of the target vehicle.

6. The lane change detection method according to claim 1, characterized in that: The step of fitting the third corner point offset ratios and the corresponding acquisition times corresponding to the plurality of historical corner point positions of the target corner point to generate a fitting curve includes: Calculating the root mean square error of the distance from the fitting point corresponding to the multiple historical corner point positions of each of the target corner points to the fitting curve; When the root mean square error is greater than a preset error threshold, it is determined that the multiple historical corner point positions of the target corner point are invalid.

7. The lane change detection method according to claim 1, characterized in that: The lane change detection result includes whether to change lanes, and the lane change detection of the target vehicle based on at least one of the corner point offset speed and the second corner point offset ratio includes: When the corner point deviation speed is greater than a preset speed threshold, determining that the target vehicle is changing lanes; and / or When the second corner point offset ratio is greater than a second preset ratio threshold, it is determined that the target vehicle is changing lanes.

8. An electronic device, characterized in that: include: A processor, the processor is connected to a memory; a computer program is stored in the memory, and the processor executes the computer program to implement instructions of the lane change detection method according to any one of claims 1 to 7.

9. A vehicle, characterized in that: include: The electronic device as claimed in claim 8.

10. A non-volatile computer-readable storage medium containing a computer program, characterized in that: When the computer program is executed by a processor, the processor executes the lane change detection method according to any one of claims 1 to 7.

Images