A lane centerline detection method, device and storage medium

Abstract

The application discloses a lane center line detection method, equipment and a storage medium. The method comprises the following steps: acquiring a target area containing a target road; acquiring a target vehicle flow driving through each first position in the target area in a first preset historical time period; and determining a center line of each lane in the target road based on the target vehicle flow of each first position. In this way, the application can improve the accuracy of lane center line detection.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method, device and storage medium for lane centerline detection. Background Technology

[0002] In some scenarios, lane centerlines play a crucial role in subsequent tasks such as prediction and planning. For example, in refined traffic management scenarios, it is necessary to determine each lane based on lane lines or lane centerlines in order to monitor the operating status of vehicles in each lane. However, some road conditions are complex, making it difficult to obtain accurate road lane information. The accuracy of lane centerline detection results often affects subsequent tasks such as prediction and planning.

[0003] Therefore, accurately detecting the centerline of each lane on the road is crucial for the execution of subsequent tasks. Summary of the Invention

[0004] The main technical problem addressed by this application is to provide a lane centerline detection method, device, and storage medium that can improve the accuracy of lane centerline detection.

[0005] To solve the above-mentioned technical problems, one technical solution adopted in this application is: to provide a lane centerline detection method, the method comprising: acquiring a target area containing a target road; acquiring the target traffic flow of vehicles passing through each first position in the target area within a first preset historical time period; and determining the centerline of each lane in the target road based on the target traffic flow of each first position.

[0006] The process of determining the centerline of each lane in the target road based on the target traffic flow at each first location includes: selecting several target attention locations from each first location based on the target traffic flow at each first location; and determining the centerline of each lane based on the several target attention locations.

[0007] Specifically, based on the target traffic flow at each first location, several target attention locations are selected from each first location, including: obtaining a preset statistical value of the target traffic flow at each first location as a reference traffic flow; comparing the target traffic flow at each first location with the reference traffic flow to obtain the comparison result of each first location; and finding at least one first location whose comparison result satisfies a preset size relationship as a target attention location.

[0008] Wherein, the preset statistical value is a statistical value characterizing the central tendency of the target traffic flow at each first location; and / or, the preset size relationship is that the target traffic flow is greater than or equal to the reference traffic flow; and / or, finding at least one first location whose comparison result satisfies the preset size relationship as the target attention location includes: generating a mask image of the target area using the comparison result of each first location, setting the value of the effective pixel in the mask image to a first value, setting the value of the remaining pixel to a second value, and the effective pixel being the pixel corresponding to the first location whose comparison result satisfies the preset size relationship; and determining the first location corresponding to the effective pixel as the target attention location.

[0009] The process of determining the centerline of each lane based on several target attention locations includes: determining several lane center points based on the distribution of several target attention locations in the target area; and using the several lane center points to obtain the centerline of each lane.

[0010] Specifically, based on the distribution of several target attention locations in the target area, several lane center points are determined, including: selecting several boundary locations from each target attention location according to the adjacent location attributes, wherein the adjacent location attribute is whether the adjacent location of the target attention location is a target attention location or an invalid first location, and an invalid first location is a first location where the target traffic flow and the reference traffic flow do not meet a preset size relationship, and the reference traffic flow is obtained by statistically analyzing the target traffic flow of each first location; using several boundary locations, at least one group of valid boundary locations is obtained, and a lane center point is determined from each group of valid boundary locations.

[0011] Specifically, based on the adjacent position attributes of each target attention position, several boundary positions are selected from each target attention position, including: each target attention position is taken as the current attention position; in response to the adjacent position attribute of the current attention position being the first attribute, the current attention position is determined as the left boundary position, the left adjacent position of the current attention position with the first attribute being the invalid first position, and the right adjacent position being the target attention position; in response to the adjacent position attribute of the current attention position being the second attribute, the current attention position is determined as the right boundary position, the left adjacent position of the current attention position with the second attribute being the target attention position, and the right adjacent position being the invalid first position.

[0012] The process includes several boundary locations, including left boundary locations and right boundary locations. At least one set of valid boundary locations is obtained using several boundary locations. The process also includes: selecting valid left boundary locations and valid right boundary locations based on the target traffic flow of each boundary location and its neighboring locations; and forming at least one set of valid boundary locations using the valid left boundary locations and valid right boundary locations, with each set of valid boundary locations including a valid left boundary location and a valid right boundary location.

[0013] Specifically, based on the target traffic flow at each boundary location and its neighboring locations, the effective left boundary location and effective right boundary location are selected, including: treating each boundary location as a target location; in response to the target location being a left boundary location and the target location having a first relational location, determining the target location as a valid left boundary location, wherein the first relational location is one of a plurality of consecutive target attention locations to the right of the target location, and the target traffic flow at the first relational location is greater than or equal to the target traffic flow at the target location; in response to the target location being a right boundary location and the target location having a second relational location, determining the target location as a valid right boundary location, wherein the second relational location is one of a plurality of consecutive target attention locations to the left of the target location, and the target traffic flow at the second relational location is greater than or equal to the target traffic flow at the target location.

[0014] Specifically, by utilizing the effective left boundary position and the effective right boundary position, at least one group of effective boundary positions is formed, including: taking the effective left boundary position and the effective right boundary position that satisfy a preset positional relationship as a group of boundary positions, wherein the preset positional relationship is that the first position between the effective left boundary position and the effective right boundary position are both target interest positions, or the effective left boundary position is the nearest effective left boundary position of the effective right boundary position and the effective right boundary position is the nearest effective right boundary position of the effective left boundary position; taking each group of boundary positions as a group of effective boundary positions, or taking the group of boundary positions whose internal distance is less than or equal to a preset threshold as a group of effective boundary positions, wherein the internal distance of the group of boundary positions is the distance between the effective right boundary position and the effective left boundary position in the group of boundary positions.

[0015] Specifically, determining a lane center point from each group of valid boundary positions includes: for each group of valid boundary positions, forming a position sequence corresponding to the valid boundary position group by combining each valid boundary position in the group with the first position located between the valid boundary position groups; finding the second and third positions from the position sequence corresponding to the valid boundary position group; and determining the lane center point corresponding to the valid boundary position group between the second and third positions. The second position is the first position from left to right in the position sequence that meets the preset traffic flow requirement, and the third position is the position located to the right of the second position and the last position from left to right in the position sequence that meets the preset traffic flow requirement. The preset traffic flow requirement is that the target traffic flow of the position is greater than or equal to the target traffic flow of its right neighbor position.

[0016] The process involves using several lane center points to obtain the center line of each lane, including: determining the lane to which each lane center point belongs based on its position; and performing linear fitting on the lane center points belonging to the same lane to obtain the lane center line.

[0017] After determining the centerline of each lane in the target road, the method further includes: acquiring first target driving data of several first target vehicles in each lane; determining the target driving status and driving direction of the first target vehicles based on the first target driving data; and updating the lane target status road segment of each lane based on the target driving status and driving direction of each first target vehicle, wherein the lane target status road segment includes at least one of congested road segment, slow-moving road segment, and smooth road segment.

[0018] The target driving state includes one of the following: a stopped state, a slow-moving state, and a normal driving state. The target driving data is collected by target sensors installed on the target road, and includes driving speed and the distance between the first target vehicle and the target sensor. Based on the target driving state and driving direction of each first target vehicle, the target state road segment of each lane is updated, including: identifying several second target vehicles currently in a slow-moving road segment, and selecting a third target vehicle from the several second target vehicles that meets the preset state requirements. The preset state requirements are that the driving state of the third target vehicle is not a normal driving state, and the number of frames of the third target vehicle disappearing within a second preset historical time period is greater than or equal to a second preset threshold. Based on the target driving data and driving direction of each third target vehicle, it is determined whether each third target vehicle is in a stopped state. In response to being in a stopped state, the stopped road segment and the slow-moving road segment are updated using the position of the third target vehicle in the stopped state.

[0019] The target driving data includes at least one of speed, acceleration, and jerk. Based on the target driving data and driving direction of the third target vehicle, it is determined whether each third target vehicle is in a stopped state, including: in response to the third target vehicle's driving direction being a first direction away from the target sensor, and its speed being greater than a first constant, its acceleration being less than or equal to the first constant, and its jerk value being less than a second constant, it is determined to be in a stopped state; in response to the third target vehicle's driving direction being a first direction away from the target sensor, and the third target vehicle being in the lane, and its speed being greater than the first constant, its acceleration being less than or equal to the first constant, and its jerk value being greater than or equal to the second constant and less than or equal to the third constant, it is determined to be in a stopped state; in response to the third target vehicle's driving direction being a second direction closer to the target sensor, and its speed being less than or equal to the first constant, its acceleration being greater than the first constant, and its jerk value being greater than the second constant, it is determined to be in a stopped state; in response to the third target vehicle's driving direction being a second direction closer to the target sensor, and the third target vehicle being in the lane, and its speed being greater than the first constant, its acceleration being less than or equal to the first constant, and its jerk value being greater than or equal to the second constant and less than or equal to the third constant, it is determined to be in a stopped state.

[0020] To solve the above-mentioned technical problems, another technical solution adopted in this application is: to provide an electronic device, including a memory and a processor coupled to each other, wherein the memory stores program instructions; and the processor is used to execute the program instructions stored in the memory to implement the above-mentioned method.

[0021] To solve the above-mentioned technical problems, another technical solution adopted in this application is to provide a computer-readable storage medium for storing program instructions that can be executed to implement the above-mentioned method.

[0022] The above scheme first obtains the target traffic flow at each first location within the target area during a first preset historical time period, and then determines the centerline of each lane in the target road based on the target traffic flow at each first location. Since drivers have different driving habits and their vehicles follow different trajectories within the lanes, most drivers' trajectories are close to the lane centerline. The target traffic flow at each first location in this application can represent the number of vehicles passing through that first location. Therefore, compared to methods that determine the lane centerline solely based on vehicle trajectories, this application's method of determining the lane centerline based on target traffic flow considers drivers' driving habits, thus avoiding the impact of different vehicle trajectories on lane centerline detection and improving the accuracy of lane centerline detection. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating an embodiment of the lane centerline detection method provided in this application;

[0024] Figure 2 yes Figure 1 The flowchart of step S13 shown is a schematic diagram of one embodiment.

[0025] Figure 3 yes Figure 2 The flowchart of step S21 shown is a schematic diagram of one embodiment;

[0026] Figure 4 yes Figure 2 The flowchart of step S22 shown is a schematic diagram of one embodiment;

[0027] Figure 5 yes Figure 4 The flowchart of step S41 shown is a schematic diagram of one embodiment.

[0028] Figure 6 yes Figure 5 The flowchart of step S52 shown is a schematic diagram of one embodiment;

[0029] Figure 7 yes Figure 6The flowchart of step S62 shown is a schematic diagram of one embodiment.

[0030] Figure 8 yes Figure 1 A flowchart of an embodiment following step S13 is shown.

[0031] Figure 9 yes Figure 8 The flowchart of step S83 shown is a schematic diagram of one embodiment.

[0032] Figure 10 Comparison images of parking sections and slow-moving sections before and after the update;

[0033] Figure 11 This is a schematic diagram of the structure of an embodiment of the electronic device provided in this application;

[0034] Figure 12 This is a schematic diagram of the structure of the computer-readable storage medium provided in this application. Detailed Implementation

[0035] To make the purpose, technical solution and effects of this application clearer and more explicit, the following describes this application in further detail with reference to the accompanying drawings and embodiments.

[0036] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0037] It should be noted that the lane centerline detection method of this application is applicable to any road scenario requiring lane centerline detection, such as structured roads and unstructured roads. It is also applicable to any scenario requiring subsequent detection and planning based on lane centerlines, such as traffic control based on detected lane centerlines. Furthermore, this application determines the lane centerlines based on the target traffic flow at each first location within the target area over a past period, rather than based on the vehicle's driving trajectory. Determining the lane centerlines based on the target traffic flow at each first location avoids the impact of different driver trajectories on the accuracy of lane centerline detection, thus improving the accuracy of lane centerline detection.

[0038] Please see Figure 1 , Figure 1 This is a schematic flowchart of an embodiment of the lane centerline detection method provided in this application. It should be noted that if substantially the same result is obtained, this embodiment is not necessarily identical. Figure 1 The illustrated process sequence is limited. For example... Figure 1 As shown, this embodiment includes:

[0039] S11: Get the target area containing the target road.

[0040] This embodiment is used to determine the centerline of each lane in the target road based on the target traffic flow passing through each first location in the target area within a first preset historical time period.

[0041] The target area is the area containing the target road, which is the road for which lane centerline detection is to be performed. The target road can be any type of road, such as a structured road or an unstructured road, and can include a crossroads or a T-junction. The target road includes at least one lane.

[0042] In one embodiment, the target area is the region where a target sensor installed on the target road can collect vehicle data. The target sensor may be, but is not limited to, millimeter-wave radar, and may also be lidar or a camera, etc.

[0043] S12: Obtain the target traffic flow of vehicles passing through the target area at each first location within the first preset historical time period.

[0044] In this embodiment, the target area includes multiple first locations, and the multiple first locations constitute the target area. For each first location, the target traffic flow is the number of vehicles that pass through the first location within a first preset historical time period.

[0045] The specific first preset historical segment can be determined based on the traffic flow or detection accuracy on the road in the actual scenario. It is understood that the more vehicles passing through the target road are collected, the more accurate the detection of the lane center line will be.

[0046] The first preset historical time period includes multiple historical time points. In one embodiment, the first location passed by vehicles at each historical time point can be obtained first. Then, for each first location, the passing situation of vehicles at each historical time point in the first preset historical time period is statistically analyzed to obtain the target traffic flow passing through the first location in the first preset historical time period.

[0047] In this embodiment, the target traffic flow at each first location can be obtained by collecting and statistically analyzing information from the target sensor, or by first acquiring the frame information collected by the target sensor and then using the frame information to statistically analyze the target traffic flow at each first location.

[0048] It should be noted that, in order to reduce the influence of vehicle trajectory and other factors on each first position, in one embodiment, the current first position can be filtered using several neighboring first positions to obtain a filtered first position. For example, the average of the coordinates of the current first position and several neighboring first positions can be used as the filtered current first position.

[0049] S13: Based on the target traffic flow at each first location, determine the centerline of each lane in the target road.

[0050] In one embodiment, considering that drivers travel between the two lane lines corresponding to each lane, and that most drivers' trajectories are close to the center of the two lane lines, to reduce the impact of different driving habits on lane centerline detection, based on the target traffic flow at each first location, several first locations with a high number of vehicles passing through can be selected as target focus locations. After selecting several target focus locations, the centerline of each lane is determined based on these target focus locations. The specific method for determining the centerline of each lane in the target road based on the target traffic flow at each first location is described below. Figure 2 The relevant explanations are as follows.

[0051] In another embodiment, a pre-trained model or related algorithm can be used to detect the target traffic flow at each first location and determine the centerline of each lane in the target road.

[0052] The above scheme first obtains the target traffic flow at each first location within the target area during a first preset historical time period, and then determines the centerline of each lane in the target road based on the target traffic flow at each first location. Since drivers have different driving habits and their vehicles follow different trajectories within the lanes, most drivers' trajectories are close to the lane centerline. The target traffic flow at each first location in this application can represent the number of vehicles passing through that first location. Therefore, compared to methods that determine the lane centerline solely based on vehicle trajectories, this application's method of determining the lane centerline based on target traffic flow considers drivers' driving habits, thus avoiding the impact of different driver trajectories on lane centerline detection and improving the accuracy of lane centerline detection.

[0053] Please see Figure 2 , Figure 2 yes Figure 1 The flowchart of one embodiment of step S13 is shown. In this embodiment, step S13 further includes S21-S22:

[0054] S21: Based on the target traffic flow of each first position, select several target attention positions from each first position.

[0055] It should be noted that most vehicles travel close to the center line of the lane. Therefore, the closer the first position in the target area is to the center line of the lane, the greater the traffic flow passing through that first position. Therefore, in order to accurately determine the center line of the lane, this embodiment can first select the first positions with more vehicles passing through as several target attention positions, and then determine the center line of each lane based on the selected target attention positions.

[0056] In one embodiment, a reference traffic flow can be directly set to determine the target traffic flow at each first location. The target traffic flow at each first location is compared with the reference traffic flow to obtain the comparison results. Then, based on the comparison results, the first location with a target traffic flow greater than or equal to the reference traffic flow is selected as the target location of interest.

[0057] In another embodiment, the target traffic flow at each first location can be pre-statistically calculated to obtain a pre-statistical value. This pre-statistical value can then be used as a reference traffic flow for determining the amount of target traffic flow at each first location. The target traffic flow at each first location can then be compared with the reference traffic flow to obtain comparison results. Based on the comparison results, the first location with a target traffic flow greater than or equal to the reference traffic flow can be selected as the target location of interest.

[0058] Specifically, please refer to Figure 3 , Figure 3 yes Figure 2 The flowchart of one embodiment of step S21 is shown. In this embodiment, step 21 further includes S31-S33:

[0059] S31: Obtain the preset statistical value of the target traffic flow at each first location as a reference traffic flow.

[0060] In this embodiment, in order to clearly distinguish the size of traffic flow, the traffic flow in the target area can be divided into two parts: one part is the target attention location with a large number of vehicles passing through, and the other part is the first location with a small number of vehicles passing through (which can be considered an invalid first location).

[0061] In this embodiment, the statistical value representing the concentration trend of the target traffic flow at each first location is first used as the preset statistical value (for example, the average value of all first locations in the target area, the middle location, etc. can be used as the preset statistical value); then the preset statistical value obtained by statistics is used as the reference traffic flow, and then the target attention location with more vehicles passing by and the first location without more vehicles passing by are distinguished according to the relationship between the target traffic flow and the reference traffic flow.

[0062] S32: Compare the target traffic flow at each first position with the reference traffic flow to obtain the comparison result for each first position.

[0063] The difference between the target traffic flow and the reference traffic flow at each first position can be calculated to obtain the difference value, which is the comparison result of each first position.

[0064] S33: Find at least one first position where the comparison result satisfies the preset size relationship, and use it as the target position of interest.

[0065] After obtaining the comparison results for each first position, at least one first position whose comparison result satisfies a preset size relationship can be selected as the target attention position. The preset size relationship is that the target traffic flow is greater than or equal to the reference traffic flow. This can be understood as the selected target attention positions having target traffic flows greater than or equal to the reference traffic flow; that is, the target attention positions are locations with a high volume of vehicles passing through.

[0066] In one specific implementation, to clearly distinguish the magnitude of traffic flow and make the distinction more intuitive, a mask image corresponding to the target area can be generated using the comparison results of the target traffic flow and the reference traffic flow at each first location. The mask image includes two types of pixels: valid pixels and all other pixels. Valid pixels in the mask image are those corresponding to the first location where the comparison result satisfies a preset size relationship (target traffic flow is greater than or equal to reference traffic flow). In this embodiment, the first location corresponding to the valid pixels is determined as the target interest location. To intuitively distinguish the magnitude of traffic flow in the target area, the values ​​of all valid pixels in the mask image can be set to a first value, and the values ​​of all other pixels can be set to a second value, with the first and second values ​​being different. For example, the first value is 1, and the second value is 0. It can be understood that, through the values ​​in the mask image of this embodiment, it is intuitive to distinguish that the first value corresponds to the target interest location with more vehicles passing by, and the second value corresponds to the first location with fewer vehicles passing by.

[0067] S22: Determine the centerline of each lane based on several target locations of interest.

[0068] In this embodiment, several lane center points can be determined based on the distribution of several target interest locations in the target area, and then the center lines of each lane can be obtained using these lane center points.

[0069] In one embodiment, please refer to Figure 4 , Figure 4 yes Figure 2 The flowchart of one embodiment of step S22 is shown. In this embodiment, step S22 further includes:

[0070] S41: Based on the distribution of several target interest locations in the target area, determine several lane center points.

[0071] As can be seen from the above, the target attention location is a location with high traffic volume, and the corresponding target location is a location close to the lane center line. Therefore, in this embodiment, the lane center point belonging to the lane center line can be determined based on the distribution of several target attention locations in the target area.

[0072] In this embodiment, several boundary positions can be selected from each target interest position based on the adjacent position attributes of each target interest position. Then, at least one group of effective boundary positions can be obtained using the several boundary positions. Finally, a lane center point can be determined from each group of effective boundary positions.

[0073] Among them, the adjacent position attribute determines whether the adjacent position of the target attention position is the target attention position or an invalid first position. An invalid first position is the first position where the target traffic flow and the reference traffic flow do not meet the preset size relationship (i.e., the target traffic flow is less than the reference traffic flow). For details on obtaining the reference traffic flow, please refer to the explanation in step S31 above.

[0074] In one specific embodiment, please refer to Figure 5 , Figure 5 yes Figure 4 The flowchart shown is a schematic diagram of an embodiment of step S41. In this embodiment, step S41, based on the distribution of several target interest locations in the target area, determines several lane center points, and further includes:

[0075] S51: Select several boundary locations from the locations of interest of each target based on the adjacent location attributes of each target location.

[0076] In this embodiment, during the process of selecting several boundary positions from each target position of interest based on the adjacent position attributes, each target position of interest can be used as the current position of interest. This facilitates traversing the adjacent position attributes of each target position of interest and determining the boundary position of each target position of interest based on these adjacent position attributes. The boundary positions include left and right boundary positions; that is, this embodiment can determine whether the current position of interest is a left or right boundary position based on the adjacent position attributes of each current position of interest.

[0077] Specifically, based on the adjacent location attributes of each target location of interest, several boundary locations are selected from each target location of interest, including the following steps:

[0078] First, each target's current focus location is designated as the current focus location.

[0079] Second, in response to the adjacent position attribute of the current position of interest being the first attribute, the current position of interest is determined as the left boundary position, the left adjacent position of the current position of interest being an invalid first position, and the right adjacent position being the target position of interest.

[0080] Specifically, if the left adjacent position of the currently focused position is an invalid first position and the right adjacent position is the target focused position, then the corresponding adjacent position attribute can be determined as the first attribute, and the currently focused position can be determined as the left boundary position. In other words, if the target traffic flow of the left adjacent position of the currently focused position is less than the reference traffic flow, and the target traffic flow of the right adjacent position is greater than or equal to the reference traffic flow, then the currently focused position is determined as the left boundary position.

[0081] Third, in response to the adjacent position attribute of the current position of interest being the second attribute, the current position of interest is determined as the right boundary position, the left adjacent position of the current position of interest is the target position of interest, and the right adjacent position is the invalid first position.

[0082] If, in the adjacent position attributes of the currently focused position, the left adjacent position is the first target position and the right adjacent position is an invalid focus position, the corresponding adjacent position attribute can be determined as the second attribute, and the currently focused position can be determined as the right boundary position. In other words, if the target traffic flow at the left adjacent position of the currently focused position is greater than or equal to the reference traffic flow, and the target traffic flow at the right adjacent position is less than the reference traffic flow, then the currently focused position is determined as the right boundary position.

[0083] The selection of several boundary positions from each target position of interest includes two cases: First, if the left adjacent position of the current position of interest is an invalid first position and the right adjacent position is a target position of interest, then the current position of interest is determined as the left boundary position; Second, if the left adjacent position of the current position of interest is a target first position and the right adjacent position is an invalid position of interest, then the current position of interest is determined as the right boundary position. In this embodiment, each boundary position can be determined by judging the case to which each current position of interest belongs.

[0084] S52: Using several boundary positions, obtain at least one set of valid boundary position groups, and determine the center point of a lane from each set of valid boundary position groups.

[0085] It should be noted that the boundary positions selected in step S51 are very likely to be located on both sides of the center line of a lane and in the middle of the two lane lines corresponding to that lane. However, in order to more accurately determine the positions located on both sides of the center line of a lane and in the middle of the two lane lines corresponding to that lane, at least one group of valid boundary positions can be selected from the boundary positions, and then the center point of a lane can be determined from each group of valid boundary positions.

[0086] It should be noted that the selection of boundary positions mentioned above essentially refers to selecting the boundary positions of each point on the center line of each lane. As mentioned above, the closer to the lane center line, the greater the corresponding traffic flow should be. If the traffic flow between boundary positions (between the left and right boundary positions) is greater than or equal to the traffic flow at the boundary position itself, it indicates that the boundary position meets the requirements for selection. To distinguish it from boundary positions that do not meet the requirements, the boundary position that meets the requirements is defined as a valid boundary position. Based on this, in one embodiment, valid left and valid right boundary positions can be selected first based on the target traffic flow of each boundary position and its neighboring positions. Then, at least one group of valid boundary positions can be formed using the valid left and valid right boundary positions, wherein each group of valid boundary positions includes one valid left boundary position and one valid right boundary position.

[0087] Specifically, please refer to Figure 6 , Figure 6 yes Figure 5 The flowchart of step S52 in one embodiment is shown. In this embodiment, step S52 uses several boundary positions to obtain at least one set of effective boundary position groups, and determines a lane center point from each set of effective boundary position groups, including:

[0088] S61: Based on the target traffic flow at each boundary location and the neighboring locations of the boundary locations, select the effective left boundary location and the effective right boundary location.

[0089] In this embodiment, during the process of selecting the effective left boundary position and the effective right boundary position based on the target traffic flow of each boundary position and the neighboring positions of the boundary positions, each boundary position can be used as the target position to facilitate traversing the target traffic flow of the neighboring positions of each target position and selecting the effective left boundary position and the effective right boundary position from each boundary position.

[0090] Specifically, this embodiment includes the following steps:

[0091] First, each boundary position is taken as the target position.

[0092] Second, in response to the target location being a left boundary location and the target location having a first relational location, the target location is determined to be a valid left boundary location. The first relational location is one of multiple consecutive target attention locations to the right of the target location, and the target traffic flow at the first relational location is greater than or equal to the target traffic flow at the target location.

[0093] In other words, for each target location, if the target location is a left boundary location, and among the multiple consecutive target attention locations to the right of the target location (left boundary location), there is a location where the target traffic flow is greater than or equal to the target traffic flow of the target location (left boundary location), it indicates that the target location has a first relationship location, and the target location is determined to be a valid left boundary location.

[0094] Third, in response to the target location being a right boundary location and the existence of a second relational location, the target location is determined to be a valid right boundary location. The second relational location is one of multiple consecutive target interest locations to the left of the target location, and the target traffic flow at the second relational location is greater than or equal to the target traffic flow at the target location.

[0095] In other words, for each target location, if the target location is a right boundary location, and among the multiple consecutive target attention locations to the left of the target location (right boundary location), there is a location where the target traffic flow is greater than or equal to the target traffic flow of the target location (right boundary location), it indicates that the target location has a second relationship location, and the target location is determined to be a valid right boundary location.

[0096] S62: Using the valid left boundary position and the valid right boundary position, form at least one group of valid boundary positions, wherein each group of valid boundary positions includes a valid left boundary position and a valid right boundary position.

[0097] In this embodiment, after determining each valid left boundary position and each valid right boundary position, the valid left boundary positions and valid right boundary positions that satisfy the preset positional relationship are grouped into a boundary position group, and then a valid boundary position group is selected from the boundary position group. In this embodiment, the left and right valid boundary positions in the selected valid boundary position group correspond to the left and right boundary positions on a lane.

[0098] In one embodiment, the preset positional relationship is that the first position between the effective left boundary position and the effective right boundary position is the target position of interest, that is, the first position between the effective left boundary position and the effective right boundary position of each group of boundary positions is the target position of interest.

[0099] In another embodiment, the preset positional relationship is: the effective left boundary position is the nearest effective left boundary position to the effective right boundary position, and the effective right boundary position is the nearest effective right boundary position to the effective left boundary position. That is, the effective left boundary position of each group of boundary positions is the nearest effective left boundary position to the effective right boundary position, and the effective right boundary position is the nearest effective right boundary position to the effective left boundary position.

[0100] In this embodiment, the effective left and right boundary positions of the selected effective boundary position groups are boundary positions on a single lane. In one embodiment, after determining each group of boundary positions, each group of boundary positions can be used as an effective boundary position group. In another embodiment, considering scenarios where target vehicles may change lanes on the target road, for example, a large number of vehicles changing lanes on a certain lane line between two lanes, the target attention position passed by a large number of vehicles may be a position on the lane line between the two lanes. In this case, the internal distance of the boundary position groups corresponding to the two lanes will be significantly greater than the internal distance of the boundary position groups corresponding to one lane. Therefore, in this embodiment, in order to ensure that the effective left and right boundary positions of the selected effective boundary position groups are boundary positions on a single lane, boundary position groups whose internal distance is less than or equal to a preset threshold can be used as effective boundary position groups. The internal distance of a boundary position group is the distance between the effective right boundary position and the effective left boundary position in the boundary position group.

[0101] If the internal distance of the boundary position group is greater than the preset threshold, it can be determined that the target vehicle is changing lanes on the lane line in the middle of the two lanes. In this case, if it is not possible to determine the center line of the two lanes based on several target attention positions, the center line of each lane in the two lanes can be determined based on the positional relationship between the two lanes and the lane with the determined center line, and the positional relationship between the determined lane center line and the two lanes.

[0102] It should be noted that, since the center lines of the lanes are parallel, after determining the center line of a certain lane, the determined center line can be translated according to the approximate width of the lane to obtain the center lines of each lane in the two lanes.

[0103] In this embodiment, after determining each group of effective boundary positions, a lane center point can be determined from each group of effective boundary positions.

[0104] In one embodiment, each effective boundary position in each effective boundary position group and the position corresponding to the maximum target traffic flow located between effective boundary position groups can be used as a center point of the lane, or the midpoint between the positions corresponding to the maximum target traffic flow and the second largest target traffic flow closest to the maximum target traffic flow can be used as a center point of the lane.

[0105] In another embodiment, a second and a third position can be selected from each effective boundary position in the effective boundary position group and the target traffic flow at the first position located between the effective boundary position groups, and then the lane center point corresponding to the effective boundary position group can be determined from the second and third positions.

[0106] Specifically, please refer to Figure 7 , Figure 7 yes Figure 6 The flowchart of one embodiment of step S62 is shown. In this embodiment, step S62 determines a lane center point from the effective boundary position group, including:

[0107] S71: For each group of valid boundary positions, the valid boundary positions in the group and the first position located between the valid boundary positions are combined to form the position sequence corresponding to the valid boundary position group. The second and third positions are found from the position sequence corresponding to the valid boundary position group.

[0108] In this embodiment, for each group of valid boundary positions, the positions are sorted according to each valid boundary position in the group and the first position located between the valid boundary positions to form a position sequence corresponding to the valid boundary position group. Then, the second and third positions used to determine the lane center point are found from the position sequence corresponding to the valid boundary position group.

[0109] S72: Determine the lane center point corresponding to the effective boundary position group between the second and third positions.

[0110] The second position is the first position from left to right in the position sequence (the position sequence corresponding to the effective boundary position group) that meets the preset traffic flow requirement. The third position is the position located to the right of the second position and the last position from left to right in the position sequence that meets the preset traffic flow requirement. The preset traffic flow requirement is that the target traffic flow of the position is greater than or equal to the target traffic flow of its right neighbor position.

[0111] In other words, the second position is the first position in the position sequence whose target traffic flow is greater than or equal to the target traffic flow of its right neighbor position and is closest to the effective left boundary; the third position is located to the right of the second position and is the first position farthest from the second position among the positions whose target traffic flow is greater than or equal to the target traffic flow of its right neighbor position.

[0112] In this embodiment, after determining the second and third positions, the midpoint between the second and third positions can be used as the lane center point, or the first position with the largest target traffic flow between the second and third positions can be selected as the lane center point. Of course, it is also possible to determine the two first positions with the largest and second largest target traffic flow between the second and third positions, and then determine the lane center point corresponding to the effective boundary position group based on these two first positions. The specific determination can be made according to the actual detection effect, and no specific limitation is made here.

[0113] S42: Using the center points of several lanes, obtain the center line of each lane.

[0114] In this embodiment, after determining several lane center points, the lane to which each lane center point belongs is determined based on the position of each lane center point. Then, the lane center points belonging to the same lane are linearly fitted to obtain the center line of the lane.

[0115] It should be noted that among several lane center points, the lane to which each lane center point belongs can be determined based on the positional relationship of the center points in the lateral direction. For example, by sorting the positions of the center points in order, a position sequence of the center points is obtained, where each point in the position sequence corresponds to a lane.

[0116] After determining the lane to which the center point of each lane belongs, the center point with the closest longitudinal distance is taken as the center point belonging to the same lane, so as to determine multiple center points belonging to each lane. Then, the center points of the lanes belonging to the same lane are linearly fitted to obtain the center line of the lane.

[0117] In some embodiments, after determining the centerline of each lane in the target road, traffic congestion analysis can also be performed using the determined centerlines of each lane.

[0118] For example, please refer to Figure 8 , Figure 8 yes Figure 1The flowchart below shows an embodiment following step S13. It should be noted that if substantially the same result is achieved, this embodiment does not necessarily follow that path. Figure 8 The illustrated process sequence is limited. For example... Figure 8 As shown, this embodiment includes:

[0119] S81: Obtain target driving data of several first target vehicles in each lane.

[0120] In this embodiment, after determining the lane lines of each lane, the lane where the first target vehicle is located can be determined based on the driving trajectory information of the first target vehicle. For example, if the driving trajectory of the first target vehicle is closest to the center line A of lane A, then it can be determined that the first target vehicle is in the lane corresponding to the center line A. The driving trajectory of the first target vehicle can be collected by a target sensor installed on the target road.

[0121] In addition, the target vehicle's driving data can also be collected by target sensors installed on the target road. These target sensors can be, but are not limited to, millimeter-wave radar, lidar, cameras, etc. They can be one type of sensor or multiple types, and the number of target sensors can be determined based on the actual situation.

[0122] In one embodiment, the target driving data includes driving speed and the distance between the first target vehicle and the target sensor; in other embodiments, the target driving data may also include at least one of driving speed, acceleration, and jerk.

[0123] S82: Based on the target driving data, determine the target driving status and driving direction of the first target vehicle.

[0124] The target driving state includes one of the following: stopped, slow-moving, and normal driving. The driving direction includes the first direction (coming direction) and the second direction (going direction).

[0125] In one embodiment, the target's driving state and direction can be determined directly based on the speed of the first target vehicle and the distance between the first target vehicle and the target sensor.

[0126] For example, when the distance between the first target vehicle and the target sensor increases, the direction of travel of the first target vehicle is determined as the first direction (outbound); when the distance between the first target vehicle and the target sensor decreases, the direction of travel of the first target vehicle is determined as the second direction (inbound). Furthermore, when the vehicle's speed is 0, or when the distance between the first target vehicle and the target sensor does not change, the target's driving state can be determined as a stopped state; when the first target vehicle's speed exceeds a threshold, the target's driving state can be determined as a normal driving state.

[0127] Of course, in other embodiments, considering the complexity of vehicle driving scenarios, the specific target driving state can also be determined based on the actual situation, combining the vehicle's acceleration and / or jerk. For example, when the first target vehicle passes through an intersection, the traffic light is red, the first target vehicle is still some distance away from the vehicle in front, and the first target vehicle does not stop but continues to drive at a very low speed (close to the speed of stopping). In this case, the speed can be combined with acceleration and jerk to determine whether the first target vehicle should stop.

[0128] S83: Based on the target driving status and driving direction of each first target vehicle, update the lane target status road segment of each lane. The lane target status road segment includes at least one of congested road segment, slow-moving road segment, and smooth road segment.

[0129] In this embodiment, if the target vehicle is determined to be in a parked state, its location can be determined, and the congested sections of the corresponding driving lane can be updated using this location. If the target vehicle is in a normal driving state, its location can be determined, and the information on unobstructed sections can be updated.

[0130] It should be noted that if the target vehicle's driving state is determined to be slow-moving based solely on its speed, and the slow-moving road segment is updated accordingly, this segment may include some congested sections. In such cases, further analysis of the second target vehicle within the slow-moving segment is needed to determine whether it is currently stopped. Using the example above: if the second target vehicle is approaching an intersection when the traffic light is red, and it is some distance from the vehicle in front, and is not stopping but continuing at a very low speed (close to stopping speed), then speed combined with acceleration and jerk can be used to determine whether the second target vehicle intends to stop.

[0131] Specifically, please refer to Figure 9 , Figure 9 yes Figure 8 The flowchart shown is a schematic diagram of one embodiment of step S83. It should be noted that if substantially the same result is achieved, this embodiment does not necessarily follow the same pattern. Figure 9 The illustrated process sequence is limited. For example... Figure 9 As shown, this embodiment includes:

[0132] S91: Identify several second target vehicles currently in a slow-moving section of road, and select a third target vehicle from among these second target vehicles that meets the preset state requirements.

[0133] In this embodiment, the preset state requirement is that the third target vehicle's driving state is not normal, and the number of frames in which the third target vehicle disappears within a second preset historical time period is greater than or equal to a second preset threshold. Step S91 is to select vehicles that are likely in a stopped state from among several second target vehicles currently in a slow-moving section. It is understood that normally moving vehicles generally have higher speeds and therefore cannot be stopped. Furthermore, if the number of frames in which the second target vehicle disappears is less than or equal to the second preset threshold, it indicates that the target is obstructed in some frames, and the target sensor did not collect vehicle information in those frames. The corresponding result of detecting that it is in a slow-moving state may be inaccurate, so it is necessary to re-evaluate such targets. Therefore, this embodiment sets filtering conditions (preset state requirements) to select third target vehicles that meet the preset state requirements for subsequent stopping judgment.

[0134] S92: Based on the target driving data and driving direction of each third target vehicle, determine whether each third target vehicle is in a parked state.

[0135] In this embodiment, the target driving data includes at least one of speed, acceleration, and jerk. To more accurately determine whether each third target vehicle is in a stopped state, the target driving data can include speed, acceleration, and jerk, so that the stopping state of each third target vehicle can be determined by combining speed, acceleration, and jerk.

[0136] Specifically, in one embodiment, if the third target vehicle is traveling in a first direction away from the target sensor, and its speed is greater than a first constant, its acceleration is less than or equal to the first constant, and its jerk value is less than a second constant, then it is determined to be in a parking state. Specifically, it can be determined to be in a parking state.

[0137] If the third target vehicle is traveling in a direction away from the target sensor, and the third target vehicle is in the lane, and its speed is greater than a first constant, its acceleration is less than or equal to a first constant, and its jerk value is greater than or equal to a second constant and less than or equal to a third constant, then it is determined to be in a parking state. Specifically, it can be determined to be in a parking state.

[0138] If the third target vehicle is traveling in a second direction closer to the target sensor, and its speed is less than or equal to a first constant, its acceleration is greater than a first constant, and its jerk value is greater than a second constant, then it is determined to be in a parking state. Specifically, it can be determined to be in an oncoming parking state.

[0139] If the third target vehicle is traveling in a direction closer to the target sensor in the second direction, and the third target vehicle is in the lane, and its speed is greater than the first constant, its acceleration is less than or equal to the first constant, and its jerk value is greater than or equal to the second constant and less than or equal to the third constant, then it is determined to be in a parking state. Specifically, it can be determined to be in an oncoming parking state.

[0140] The values ​​of the first, second, and third constants can be determined according to the actual situation. For example, if the target sensor is a millimeter-wave radar, the second and third constants can be set according to the frame rate of the millimeter-wave radar. In a specific embodiment, the first constant is 0. When the target sensor is a millimeter-wave radar and the frame rate is 10, the second constant can be set to 0.8-1.5, and the third constant can be set to 0-0.5.

[0141] It should be noted that the parameters such as driving speed, velocity, or acceleration mentioned in this embodiment refer to the driving parameters of the first or third target vehicle relative to the target sensor. For example, the speed of the third target vehicle in this text refers to the driving speed of the third target vehicle relative to the target sensor, not the driving speed of the first or third target vehicle itself. Taking the third target vehicle as an example, when the third target vehicle is traveling in a first direction away from the target sensor, if the third target vehicle is getting farther and farther away from the target sensor, then the speed of the third target vehicle is greater than a first constant. When the third target vehicle is traveling in a second direction closer to the target sensor, if the third target vehicle is getting closer and closer to the target sensor, then the speed of the third target vehicle is less than or equal to the first constant.

[0142] S93: In response to a stopped state, update the stopped road segment and slow-moving road segment using the location of the third target vehicle that is in a stopped state.

[0143] If the third target vehicle is determined to be parked, then the location of the parked third target vehicle is used to update the parking segment and the slow-moving segment, resulting in the updated parking segment and slow-moving segment. For example... Figure 10 As shown, Figure 10 Comparison images of parking sections and slow-moving sections before and after the update. Figure 10 Some slow-moving sections of the road will be converted into parking areas.

[0144] Please see Figure 11 , Figure 11 This is a schematic diagram of an embodiment of the electronic device provided in this application. In this embodiment, the electronic device 110 includes a processor 101 and a memory 102.

[0145] Processor 101 can also be referred to as CPU (Central Processing Unit). Processor 101 may be an integrated circuit chip with signal processing capabilities. Processor 101 can also be a general-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component. The general-purpose processor can be a microprocessor, or processor 101 can be any conventional processor 101, etc.

[0146] The memory 102 in the electronic device 110 is used to store the program instructions required for the processor 101 to run.

[0147] The processor 101 is used to execute program instructions to implement the methods provided in any of the above embodiments and any non-conflicting combinations thereof.

[0148] In one embodiment, the electronic device may be a target sensor.

[0149] Please see Figure 12 , Figure 12 This is a schematic diagram of the structure of the computer-readable storage medium provided in this application. The computer-readable storage medium 120 of this application embodiment stores program instructions 121, which, when executed, implement the methods provided in any of the above embodiments and any non-conflicting combinations. The program instructions 121 can form a program file and be stored in the computer-readable storage medium 120 in the form of a software product, so that a computer device (which may be a personal computer, server, or network device, etc.) can execute all or part of the steps of the methods of various embodiments of this application. The aforementioned computer-readable storage medium 120 includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks, or terminal devices such as computers, servers, mobile phones, and tablets.

[0150] The above scheme first obtains the target traffic flow at each first location within the target area during a first preset historical time period, and then determines the centerline of each lane in the target road based on the target traffic flow at each first location. Since drivers have different driving habits and their vehicles follow different trajectories within the lanes, most drivers' trajectories are close to the lane centerline. The target traffic flow at each first location in this application can represent the number of vehicles passing through that first location. Therefore, compared to methods that determine the lane centerline solely based on vehicle trajectories, this application's method of determining the lane centerline based on target traffic flow considers drivers' driving habits, thus avoiding the impact of different vehicle trajectories on lane centerline detection and improving the accuracy of lane centerline detection.

[0151] The description of the various embodiments above tends to emphasize the differences between the various embodiments. The similarities or similarities between them can be referred to, and for the sake of brevity, they will not be repeated here.

[0152] The above description is merely an embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for detecting lane centerlines, characterized in that, The method includes: Obtain the target area containing the target road; Obtain the target traffic flow rate of vehicles passing through each first location in the target area within a first preset historical time period; Based on the target traffic flow at each of the first locations, several target attention locations are selected from each of the first locations; the target traffic flow at each of the target attention locations is greater than or equal to the reference traffic flow; the reference traffic flow is a preset statistical value of the target traffic flow at each of the first locations; Based on the distribution of the aforementioned target attention locations in the target area, several lane center points are determined; wherein, each lane center point is determined by each group of effective boundary positions obtained using several boundary positions, and the several boundary positions are selected from each of the aforementioned target attention locations using the adjacent position attributes of each of the aforementioned target attention locations; the several boundary positions include the left boundary position and the right boundary position used to characterize the boundary position of the lane center point. Using the center points of the lanes, the center lines of each lane are obtained; Based on the target traffic flow at each of the first locations, several target locations of interest are selected from each of the first locations, including: The target traffic flow at each of the first locations is compared with the reference traffic flow to obtain the comparison result for each of the first locations; Find at least one first position whose comparison result satisfies a preset size relationship, and use it as the target focus position.

2. The method according to claim 1, characterized in that, The preset statistical value is a statistical value that characterizes the concentration trend of the target traffic flow at each of the first locations; And / or, finding at least one first position where the comparison result satisfies a preset size relationship, as the target focus position, includes: A mask image of the target region is generated using the comparison results of each of the first positions. The values ​​of the effective pixels in the mask image are set to a first value, and the values ​​of the remaining pixels are set to a second value. The effective pixels are the pixels corresponding to the first positions that satisfy a preset size relationship with the comparison results. The first position corresponding to the effective pixel is determined as the target attention position.

3. The method according to claim 1, characterized in that, The adjacent position attribute determines whether the adjacent position of the target attention position is a target attention position or an invalid first position. The invalid first position is the first position where the target traffic flow and the reference traffic flow do not satisfy a preset size relationship. The reference traffic flow is obtained by statistically analyzing the target traffic flow at each of the first positions.

4. The method according to claim 3, characterized in that, Based on the adjacent location attributes of the target location of interest, several boundary locations are selected from the target locations of interest, including: Each of the aforementioned target attention locations is designated as the current attention location; In response to the adjacent position attribute of the current position of interest being a first attribute, the current position of interest is determined to be a left boundary position, the first attribute being that the left adjacent position of the current position of interest is an invalid first position, and the right adjacent position is the target position of interest; In response to the adjacent position attribute of the current position of interest being the second attribute, the current position of interest is determined to be the right boundary position, and the second attribute is that the left adjacent position of the current position of interest is the target position of interest, and the right adjacent position is an invalid first position.

5. The method according to claim 1, characterized in that, Each set of valid boundary positions is obtained using the aforementioned boundary positions, including: Based on the target traffic flow at each of the boundary locations and the neighboring locations of the boundary locations, the effective left boundary location and the effective right boundary location are selected. Each group of valid boundary positions is formed by using the valid left boundary position and the valid right boundary position.

6. The method according to claim 5, characterized in that, The selection of effective left and right boundary positions based on the target traffic flow at each boundary position and its neighboring positions includes: Each of the aforementioned boundary positions is taken as the target position; In response to the target location being a left boundary location and the target location having a first relational location, the target location is determined to be a valid left boundary location, wherein the first relational location is one of a plurality of consecutive target attention locations to the right of the target location, and the target traffic flow at the first relational location is greater than or equal to the target traffic flow at the target location. In response to the target location being a right boundary location and the target location having a second relational location, the target location is determined to be a valid right boundary location, wherein the second relational location is one of a plurality of consecutive target attention locations to the left of the target location, and the target traffic flow at the second relational location is greater than or equal to the target traffic flow at the target location.

7. The method according to claim 5, characterized in that, The effective left boundary position and the effective right boundary position are used to form each group of effective boundary positions, including: The effective left boundary position and the effective right boundary position that satisfy the preset position relationship are taken as a group of boundary positions. The preset position relationship is that the first position between the effective left boundary position and the effective right boundary position are both target interest positions, or the effective left boundary position is the nearest effective left boundary position of the effective right boundary position and the effective right boundary position is the nearest effective right boundary position of the effective left boundary position. Each group of boundary positions is taken as the effective boundary position group, or the boundary position group whose internal distance is less than or equal to a preset threshold is taken as the effective boundary position group, and the internal distance of the boundary position group is the distance between the effective right boundary position and the effective left boundary position in the boundary position group.

8. The method according to claim 1, characterized in that, Using each set of valid boundary positions obtained from several boundary positions, determine the center point of each lane, including: For each group of valid boundary positions, the valid boundary positions in the group and the first position located between the valid boundary positions are combined to form the position sequence corresponding to the valid boundary position group. The second and third positions are then found from the position sequence corresponding to the valid boundary position group. The lane center point corresponding to the effective boundary position group is determined between the second position and the third position; Wherein, the second position is the first position from left to right in the position sequence that meets the preset traffic flow requirement, and the third position is the position located to the right of the second position and the last position from left to right in the position sequence that meets the preset traffic flow requirement. The preset traffic flow requirement is that the target traffic flow of the position is greater than or equal to the target traffic flow of its right neighbor position.

9. The method according to claim 1, characterized in that, The process of obtaining the centerline of each lane using the plurality of lane center points includes: Based on the location of each lane center point, determine the lane to which each lane center point belongs; The centerline of the lane is obtained by linearly fitting the center points of lanes belonging to the same lane.

10. The method according to claim 1, characterized in that, After determining the centerline of the lane, the method further includes: Obtain first target driving data of several first target vehicles in each of the lanes; Based on the first target driving data, the target driving status and driving direction of the first target vehicle are determined; Based on the target driving status and driving direction of each of the first target vehicles, the target lane status segment of each lane is updated, and the target lane status segment includes at least one of congested segment, slow-moving segment and smooth segment.

11. The method according to claim 10, characterized in that, The target driving state includes one of the following: stopped state, slow-moving state, and normal driving state; the target driving data is collected by a target sensor installed on the target road, and the target driving data includes the driving speed and the distance between the first target vehicle and the target sensor; The step of updating the lane target state road segment of each lane based on the target driving state and driving direction of each first target vehicle includes: Identify several second target vehicles currently in a slow-moving section of road, and select a third target vehicle from these second target vehicles that meets preset state requirements; wherein, the preset state requirements are that the driving state of the third target vehicle is not the normal driving state, and the number of disappearance frames of the third target vehicle within a second preset historical time period is greater than or equal to a second preset threshold. Based on the target driving data and driving direction of each of the third target vehicles, it is determined whether each of the third target vehicles is in a parked state; In response to a stopped state, the location of the third target vehicle that is stopped is used to update the stopped road segment and slow-moving road segment.

12. The method according to claim 11, characterized in that, The target driving data includes at least one of speed, acceleration, and jerk. The step of determining whether each of the third target vehicles is in a parked state based on the target driving data and driving direction of the third target vehicle includes: If the third target vehicle is traveling in a direction away from the target sensor, and its speed is greater than a first constant, its acceleration is less than or equal to the first constant, and its jerk value is less than a second constant, then it is determined to be in a stopped state. If the third target vehicle is traveling in a direction away from the target sensor, and the third target vehicle is in the lane, and the speed is greater than the first constant, the acceleration is less than or equal to the first constant, and the acceleration value is greater than or equal to the second constant and less than or equal to the third constant, then it is determined to be in a stopped state. If the third target vehicle is traveling in a second direction closer to the target sensor, and its speed is less than or equal to the first constant, its acceleration is greater than the first constant, and its jerk value is greater than the second constant, then it is determined to be in a stopped state. If the third target vehicle is traveling in a second direction closer to the target sensor, and the third target vehicle is within the lane, and the speed is greater than the first constant, the acceleration is less than or equal to the first constant, and the acceleration value is greater than or equal to the second constant and less than or equal to the third constant, then it is determined to be in a stopped state.

13. An electronic device, characterized in that, Including interconnected memory and processor, The memory stores program instructions; The processor is used to execute program instructions stored in the memory to implement the method according to any one of claims 1-12.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores program instructions that can be executed by a processor, the program instructions being used to implement the method according to any one of claims 1-12.

Images