Intersection waiting area identification method, device, equipment, medium and program product

Abstract

The invention provides an intersection turn waiting area identification method, device and equipment, a medium and a program product, and relates to the technical field of high-precision maps, and the method comprises the steps: obtaining the parking point information of each vehicle under a plurality of time steps according to the trajectory data of a vehicle waiting for a signal lamp in front of a signal lamp intersection, and the parking point information is used for indicating the parking position of the vehicle; according to the parking point information under each time step, determining an initial waiting area confidence coefficient of the parking position under each time step; and according to the initial waiting zone confidence coefficient of the parking position under each time step, determining the waiting zone confidence coefficient of the parking position, and when the waiting zone confidence coefficient reaches a preset confidence coefficient threshold value, determining an intersection waiting zone according to the parking position. Through the method, the problems of low coverage rate and high cost of manual and collection vehicle collection waiting areas can be effectively solved, and the efficiency is higher.

Description

Technical Field

[0001] This application relates to the technical field of high-precision maps, and particularly to a method, apparatus, device, medium, and program product for identifying a left-turn waiting area at an intersection. Background Art

[0002] A left-turn waiting area at an intersection is an area where vehicles waiting to turn left or right can temporarily stop. The left-turn waiting area can help vehicles wait safely for a turning signal or find a suitable opportunity to complete a turning maneuver without affecting the straight-through traffic.

[0003] In the process of making a high-precision map, it is necessary to identify the location of the left-turn waiting area at the intersection. Currently, the data of the left-turn waiting area at the intersection is mainly collected through manual collection or collection vehicle collection, which requires a large amount of manpower and professional collection equipment, consumes a large amount of human and material resources, and has a limited coverage area.

[0004] Therefore, there is an urgent need to propose a technical solution that can efficiently identify the left-turn waiting area at the intersection to solve the above technical problems. Summary of the Invention

[0005] This application provides a method, apparatus, device, medium, and program product for identifying a left-turn waiting area at an intersection to at least solve one of the above technical problems.

[0006] According to one aspect of this application, a method for identifying a left-turn waiting area at an intersection is provided, including: obtaining the stop point information of each vehicle at multiple time steps according to the stop vehicle trajectory data in front of the signalized intersection, where the stop point information is used to indicate the stop position of the vehicle; determining the initial left-turn waiting area confidence level of the stop position at each time step according to the stop point information at each time step; determining the left-turn waiting area confidence level of the stop position according to the initial left-turn waiting area confidence level of the stop position at each time step, and when the left-turn waiting area confidence level reaches a preset confidence level threshold, determining the left-turn waiting area at the intersection according to the stop position.

[0007] In an implementation manner, the stop point information includes the area where the vehicle's stop position belongs and the number of stops; determining the initial left-turn waiting area confidence level of the stop position at each time step according to the stop point information at each time step includes: for each time step, determining the initial left-turn waiting area confidence level of the stop position at the time step according to the area where the stop position of the same vehicle among the vehicles at the time step belongs and the number of stops, so as to obtain the initial left-turn waiting area confidence level of the stop position at each time step.

[0008] In one embodiment, determining the initial confidence level of the waiting-turn area of the parking position at the time step according to the area where the parking position of the same vehicle among the vehicles at the time step belongs and the number of parking times includes: obtaining a first vehicle number according to the number of vehicles whose parking positions among the vehicles at the time step belong to the straight section before the traffic light and the intersection and the number of parking times reaches two; obtaining a second vehicle number according to the number of vehicles with turning trajectory points among the vehicles at the time step and the number of vehicles waiting for the traffic light simultaneously with the vehicles on the straight section before the traffic light; calculating the initial confidence level of the waiting-turn area of the parking position at the time step according to the ratio between the first vehicle number and the second vehicle number.

[0009] In one embodiment, determining the confidence level of the waiting-turn area of the parking position according to the initial confidence level of the waiting-turn area of the parking position at each time step includes: obtaining the distribution curve of the initial confidence level of the waiting-turn area with respect to the time step by means of curve fitting according to the initial confidence level of the waiting-turn area of the parking position at each time step, to obtain the distribution curve of the confidence level of the waiting-turn area; when there is no inflection point in the distribution curve of the confidence level of the waiting-turn area, calculating the first mean value between the initial confidence levels of the waiting-turn area of the parking position at each time step, to obtain the confidence level of the waiting-turn area of the parking position.

[0010] In one embodiment, the method further includes: when there is an inflection point in the distribution curve of the confidence level of the waiting-turn area, calculating the second mean value between the initial confidence levels of the waiting-turn area of the parking position at each time step before the inflection point, and the third mean value between the initial confidence levels of the waiting-turn area of the parking position at each time step after the inflection point; obtaining the change information of the waiting-turn area regarding the parking position according to the second mean value and the third mean value, where the change information of the waiting-turn area is used to indicate that the waiting-turn area at the intersection changes from existing to non-existing or from non-existing to existing over time.

[0011] In one embodiment, the method further includes: for each time step, calculating the average speed between any adjacent trajectory points except the trajectory points corresponding to the straight section before the traffic light according to the trajectory point data of the same vehicle among the vehicles at the time step, and if the average speed is lower than the preset speed threshold, obtaining the displacement and the moving time difference between the adjacent trajectory points; when the displacement is less than the preset distance threshold and the moving time difference between the adjacent trajectory points reaches the preset staying time, determining that the vehicle is in the waiting-for-traffic-light state in the staying area corresponding to the adjacent trajectory points; obtaining the area where the parking position of the same vehicle among the vehicles at the time step belongs according to the straight section before the traffic light and the staying area.

[0012] In one embodiment, the waiting vehicle trajectory data is determined based on the matching result between the preset vehicle trajectory data and the road network, where the vehicle trajectory data is obtained based on a sliding time window, and the time window is used to indicate a preset time range and moves in time steps; obtaining the stop point information of each vehicle at multiple time steps according to the waiting vehicle trajectory data in front of the signalized intersection, including: grouping the waiting vehicle trajectory data according to the straight road section and the turning road section at the signalized intersection according to the waiting vehicle trajectory data in front of the signalized intersection to obtain grouped trajectory data; respectively obtaining the stop point information of the vehicles on the straight road section and in the intersection at the signalized intersection according to the grouped trajectory data.

[0013] In one embodiment, the waiting vehicle trajectory data is determined based on the matching result between the preset vehicle trajectory data and the road network, including: extracting the signalized intersection and the road section information at the signalized intersection according to the node information and road section information in the road network; filtering the trajectory matching data at non-signalized intersections in the matching result according to the road section information at the signalized intersection to obtain the waiting vehicle trajectory data.

[0014] According to a second aspect of the present application, there is provided an identification device for a waiting-turning area at an intersection, including: a first acquisition module configured to obtain the stop point information of each vehicle at multiple time steps according to the waiting vehicle trajectory data in front of the signalized intersection, where the stop point information is used to indicate the stop position of the vehicle; a first determination module configured to determine the initial waiting-turning area confidence level of the stop position at each time step according to the stop point information at each time step; an identification module configured to determine the waiting-turning area confidence level of the stop position according to the initial waiting-turning area confidence level of the stop position at each time step, and when the waiting-turning area confidence level reaches a preset confidence level threshold, determine the waiting-turning area at the intersection according to the stop position.

[0015] In one embodiment, the stop point information includes the stop position and the number of stops of the vehicle; the first determination module is specifically configured to, for each time step, determine the initial waiting-turning area confidence level of the stop position at the time step according to the area where the stop position of the same vehicle among the vehicles at the time step belongs and the number of stops, so as to obtain the initial waiting-turning area confidence level of the stop position at each time step.

[0016] In one implementation, determining the initial confidence level of the turning area for the parking position at the time step based on the area to which the parking position of the same vehicle in each vehicle belongs and the number of parking times at the time step includes: obtaining the first vehicle quantity according to the number of vehicles whose parking positions of the same vehicle in each vehicle at the time step belong to the straight section before the traffic signal and the intersection and the number of parking times reaches two; obtaining the second vehicle quantity according to the number of vehicles with turning trajectory points in each vehicle at the time step and the number of vehicles waiting for the traffic signal simultaneously with the vehicles on the straight section before the traffic signal; calculating the initial confidence level of the turning area for the parking position at the time step according to the ratio between the first vehicle quantity and the second vehicle quantity.

[0017] In one implementation, the recognition module includes: a curve fitting unit configured to obtain the initial turning area confidence level distribution with respect to the time step by curve fitting according to the initial turning area confidence level of the parking position at each time step, and obtain the turning area confidence level distribution curve; a first calculation unit configured to calculate the first mean value between the initial turning area confidence levels of the parking position at each time step when there is no inflection point in the turning area confidence level distribution curve, and obtain the turning area confidence level of the parking position.

[0018] In one implementation, the recognition module further includes: a second calculation unit configured to calculate the second mean value between the initial turning area confidence levels of the parking position at each time step before the inflection point and the third mean value between the initial turning area confidence levels of the parking position at each time step after the inflection point when there is an inflection point in the turning area confidence level distribution curve; a change information acquisition unit configured to obtain the turning area change information regarding the parking position according to the second mean value and the third mean value, where the turning area change information is used to indicate that the turning area at the intersection changes from having to not having or from not having to having over time.

[0019] In one implementation, the device further includes: a second acquisition module configured to, for each time step, calculate the average speed between any adjacent trajectory points except the trajectory points corresponding to the straight section before the traffic signal according to the trajectory point data of the same vehicle in each vehicle at the time step, and if the average speed is lower than a preset speed threshold, obtain the displacement and the moving time difference between the adjacent trajectory points; a second determination module configured to determine that the vehicle is in the waiting state for the traffic signal in the waiting area corresponding to the adjacent trajectory points when the displacement is less than a preset distance threshold and the moving time difference between the adjacent trajectory points reaches a preset staying time; a third acquisition module configured to obtain the area to which the parking position of the same vehicle in each vehicle at the time step belongs according to the straight section before the traffic signal and the waiting area.

[0020] In one embodiment, the waiting vehicle trajectory data is determined based on the matching result between the preset vehicle trajectory data and the road network, wherein the vehicle trajectory data is obtained based on a sliding time window, and the time window is used to indicate a preset time range and moves in time steps; the first acquisition module includes: a grouping unit configured to group the waiting vehicle trajectory data according to the straight road section and the turning road section at the signalized intersection based on the waiting vehicle trajectory data in front of the signalized intersection to obtain grouped trajectory data; a stop point acquisition unit configured to respectively acquire the straight road section at the signalized intersection and the stop point information of each vehicle within the intersection according to the grouped trajectory data.

[0021] In one embodiment, the first acquisition module is further configured to determine the waiting vehicle trajectory data based on the matching result between the preset vehicle trajectory data and the road network, and the first acquisition module further includes: an extraction unit configured to extract the signalized intersection and the road section information at the signalized intersection according to the node information and the road section information in the road network; a matching unit configured to filter the trajectory matching data at non-signalized intersections in the matching result according to the road section information at the signalized intersection to obtain the waiting vehicle trajectory data.

[0022] According to a third aspect of the present application, there is provided an electronic device, including: a memory and a processor; the memory stores computer execution instructions; the processor executes the computer execution instructions stored in the memory, so that the electronic device executes the intersection left-turn waiting area recognition method according to any one of the above first aspects.

[0023] According to a fourth aspect of the present application, there is provided a computer-readable storage medium storing computer execution instructions, and when the computer execution instructions are executed by a processor, they are used to implement the intersection left-turn waiting area recognition method provided in any one of the above first aspects.

[0024] According to a fifth aspect of the present application, there is provided a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the intersection left-turn waiting area recognition method provided in any one of the above first aspects.

[0025] The method, device, equipment, medium and program product for identifying the waiting-turn area at an intersection provided by this application obtain the stop point information of each vehicle at multiple time steps based on the trajectory data of waiting vehicles in front of the signalized intersection. This stop point information is used to indicate the stop position of the vehicle. Then, according to the stop point information at each time step, the initial waiting-turn area confidence level of the stop position at each time step is determined. Next, based on the initial waiting-turn area confidence level of the stop position at each time step, the waiting-turn area confidence level of the stop position is determined. When the waiting-turn area confidence level reaches the preset confidence threshold, the waiting-turn area at the intersection is determined according to this stop position. In this process, by analyzing the stop point information corresponding to the trajectory data of waiting vehicles at different time steps to obtain the initial waiting-turn area confidence level of the stop position at different time steps, thereby determining the waiting-turn area confidence level of the stop position, the efficient identification of the waiting-turn area at the intersection can be realized and the dynamic changes of the waiting-turn area at the intersection can be adapted, without relying on manual or collection vehicles to collect the waiting-turn area, effectively solving the problems of low coverage rate and high cost of collecting the waiting-turn area by manual or collection vehicles, and facilitating the national coverage and wide-range identification of waiting-turn areas of all road grades, with higher efficiency and wider applicability. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings here are incorporated into the description and form a part of this description, showing embodiments consistent with this application and used together with the description to explain the principles of this application.

[0027] Figure 1 It is a possible scenario schematic diagram provided by an embodiment of this application;

[0028] Figure 2 It is an example map without waiting-turn area information at the intersection;

[0029] Figure 3 It is a flowchart of a method for identifying the waiting-turn area at an intersection provided by an embodiment of this application;

[0030] Figure 4 It is an example diagram of the trajectory of waiting vehicles in an embodiment of this application;

[0031] Figure 5a It is one of the example diagrams of the waiting-turn area confidence level distribution curve in an embodiment of this application;

[0032] Figure 5b It is another example diagram of the waiting-turn area confidence level distribution curve in an embodiment of this application;

[0033] Figure 5c It is the third example diagram of the waiting-turn area confidence level distribution curve in an embodiment of this application;

[0034] Figure 6 It is one of the flowcharts of another method for identifying the waiting-turn area at an intersection provided by an embodiment of this application;

[0035] Figure 7a It is an example diagram of a single intersection in an embodiment of the present application;

[0036] Figure 7b It is an example diagram of a composite intersection in an embodiment of the present application;

[0037] Figure 8 It is a schematic flow diagram for extracting road segment information at a signalized intersection in an embodiment of the present application;

[0038] Figure 9 It is the second schematic flow diagram of another intersection left-turn waiting area recognition method provided in an embodiment of the present application;

[0039] Figure 10 It is a schematic structural diagram of an intersection left-turn waiting area recognition device provided in an embodiment of the present application;

[0040] Figure 11 It is a schematic structural diagram of an electronic device provided in an embodiment of the present application.

[0041] Through the above-mentioned drawings, specific embodiments of the present application have been shown, and there will be more detailed descriptions hereinafter. These drawings and textual descriptions are not intended to limit the scope of the concept of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. Detailed Description of the Embodiments

[0042] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0043] The embodiments of the present application will be explained below in combination with the application scenarios. The intersection left-turn waiting area recognition method provided in the embodiments of the present application can be applied to the application scenario of intelligent driving. More specifically, it can be applied to the application scenario of autonomous driving based on vehicle cloud computing. Exemplarily, the execution subject of the method provided in the embodiments of the present application can be a server. More specifically, for example, it is the high-precision (HD) map server of the high-precision map provider. Hereinafter, the server will be used as the execution subject of the method provided in the embodiments of the present application for introduction.

[0044] Figure 1 It is a schematic scenario diagram of an intersection left-turn waiting area recognition method provided in an embodiment of the present application, as Figure 1As shown in the figure, it includes an HD map server 110, a trajectory database 120, and a navigation (SD) map server 130. The HD map server 110 is connected to the trajectory database 120 and the SD map server 130 through a network respectively. The HD map server 110 is used to create a high-precision map and transmit the high-precision map data to the intelligent vehicle 140, and the intelligent vehicle 140 can use the high-precision map data to assist in autonomous driving. Among them, during the process of creating the high-precision map by the HD map server 110, it uses the vehicle trajectory data transmitted by the trajectory database 120 and the road network information transmitted by the SD map server to identify the intersection's left-turn waiting area and generate a high-precision map with left-turn waiting area data. Or, the HD map server 11 can directly use the waiting vehicle trajectory data transmitted by the trajectory database 120 to identify the intersection's left-turn waiting area. Optionally, the above servers can be independent physical servers, or a server cluster or distributed system composed of multiple physical servers, or cloud servers that provide basic cloud computing services such as cloud services, cloud databases, and cloud computing.

[0045] The left-turn waiting area of the intersection is of great significance to the high-precision map data. For example, in the application of calculating traffic lights using trajectory vehicles, the stop line information of the traffic light intersection is required. For intersections with a left-turn waiting area, the position of the stop line is actually the end position of the left-turn waiting area. Traditional map data usually does not contain information about the left-turn waiting area, as Figure 2 shown, where Figure 2 part (a) shows that the traditional map does not contain a left-turn waiting area, Figure 2 and part (b) shows an example of a vehicle trajectory turning in the left-turn waiting area (the left-turn waiting area is not included in the map, but exists in the actual road scene). Although the high-precision map provides more detailed intersection information (such as lane lines and traffic light positions), its practicality is limited due to high production costs, low update frequency, and limited coverage. For example, currently, by manually collecting the left-turn waiting area of intersections, professional Geographic Information System (GIS) practitioners often need to be equipped with complete high-precision map information collection equipment to collect left-turn waiting area data with higher accuracy. It is time-consuming and laborious, with high labor costs, and it is difficult to update frequently, resulting in low data freshness. Or using a collection vehicle to collect the left-turn waiting area of intersections instead of manual collection, although it reduces the labor cost to a certain extent, the collection vehicle needs to be equipped with sensors such as radar, cameras, and Real-Time Kinematic (RTK), with high hardware costs, mainly focusing on high-class roads, and its coverage is often limited, making it difficult to achieve large-scale identification of left-turn waiting areas at intersections. Therefore, there is an urgent need to provide a more efficient and reasonable method to facilitate the large-scale identification of left-turn waiting area intersections on roads, and even the shape topology of the left-turn waiting area is particularly important.

[0046] In view of this, the embodiments of the present application provide a method, device, equipment, medium and program product for identifying a waiting-turn area at an intersection. By obtaining the stop point information of each vehicle at multiple time steps based on the trajectory data of the waiting vehicles in front of the signalized intersection, the stop point information is used to indicate the stop position of the vehicle. And according to the stop point information at each time step, the initial waiting-turn area confidence level of the stop position at each time step is determined. Then, according to the initial waiting-turn area confidence level of the stop position at each time step, the waiting-turn area confidence level of the stop position is determined. When the waiting-turn area confidence level reaches the preset confidence threshold, the waiting-turn area at the intersection is determined according to the stop position. In this process, by analyzing the stop point information corresponding to the trajectory data of the waiting vehicles at different time steps to obtain the initial waiting-turn area confidence level of the stop position at different time steps, so as to determine the waiting-turn area confidence level of the stop position, the efficient identification of the waiting-turn area at the intersection can be realized and the dynamic changes of the waiting-turn area at the intersection can be adapted, without relying on manual or collection vehicle to collect the waiting-turn area, effectively solving the problems of low coverage rate and high cost of manual or collection vehicle to collect the waiting-turn area, and facilitating the wide-range identification of the waiting-turn area for national coverage and all road grades, with higher efficiency and wider applicability.

[0047] The technical solution of the present application and how the technical solution of the present application solves the above technical problems will be described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that these specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0048] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data need to comply with relevant laws, regulations and standards, and corresponding operation entrances are provided for users to choose to authorize or refuse.

[0049] Figure 3 is a schematic flowchart of a method for identifying a waiting-turn area at an intersection provided by an embodiment of the present application. As Figure 3 shown, the method may include the following steps S301-S304:

[0050] Step S301: Obtain the stop point information of each vehicle at multiple time steps based on the trajectory data of the waiting vehicles in front of the signalized intersection, and the stop point information is used to indicate the stop position of the vehicle.

[0051] In this embodiment, the trajectory data of the waiting vehicles in front of the traffic signal can be the waiting trajectory data for each vehicle. By using information such as the speed and position of the trajectory points in the waiting trajectory data of each vehicle, the parking point information of each vehicle can be determined. For example, when the speed of a certain trajectory point drops to zero (or other speed thresholds close to zero, determined according to empirical values), it is determined that this trajectory point is the parking position point. The parking point information of each vehicle can be obtained by using the determined parking position points and the number of parking times of each vehicle.

[0052] It can be understood that the waiting vehicle trajectory data is the trajectory data at the entrance of the traffic signal intersection. Therefore, it can effectively solve the interference problem of the speed dropping to zero caused by vehicle congestion. Using the above method of the speed dropping to zero, the parking position point can be determined more accurately. In some embodiments, it can further be checked whether the vehicle position remains stable for a period of time when the speed is zero, so as to improve the recognition accuracy of the parking position point.

[0053] In addition, this embodiment takes into account that the turning area at the intersection may change, such as the new construction or demolition of the turning area at the intersection, and may have different recognition results at different time steps. By using the waiting vehicle trajectory data to obtain the parking point information at multiple time steps, a more fine-grained (day, hour) recognition of the turning area at the intersection can be achieved, which can effectively solve the problem of low freshness of the turning area caused by the method of manually collecting or using a collection vehicle to collect the turning area. Among them, the time step is the unit time step, and each time step corresponds to a specific time interval, and this specific time interval can be determined based on the empirical values of those skilled in the art. For example, it is in units of 1h or other empirical values. This embodiment does not make special limitations on this.

[0054] Optionally, the waiting vehicle trajectory data can be pre-determined and stored in the server, or can be obtained by the server in real time according to the matching result between the preset vehicle trajectory data and the road network. Among them, the vehicle trajectory data can be obtained based on a sliding time window, and this time window is used to indicate a preset time range and moves at this time step. For example, the time window is 1 day and the time step is 1h.

[0055] For example, in this embodiment, the pre-set vehicle trajectory data can be vehicle trajectory data within a certain time interval (such as real-time trajectory data within one month) and within a certain regional scope (such as within a certain administrative division) prepared in advance, and can be adaptively adjusted according to the actual intersection approach area recognition requirements. The vehicle trajectory data, that is, the GPS point trajectory data of the vehicle, can include information such as timestamp, position coordinates (latitude and longitude), speed, and direction. The road network can include geometric information of roads, intersection positions, signal light positions, etc., and can be obtained from existing SD map data. Exemplarily, in the process of matching the GPS point trajectory data with the road network, first, a large amount of floating car trajectory data is received from the trajectory database. Each trajectory data usually includes information such as vehicle identification (CarID), timestamp, and position coordinates (latitude and longitude), and the received trajectory data can be grouped according to CarID, so that the trajectory data of each vehicle is aggregated together, facilitating subsequent processing and analysis. Moreover, the trajectory data within each CarID group can be sorted according to the timestamp, and data cleaning can be performed to remove outliers and noise data, improving the accuracy of the trajectory data. Then, a suitable map matching algorithm (such as the hidden Markov model, etc.) is selected to map the GPS trajectory of the vehicle to the specific road of the road network to perform map matching on the trajectory data within each CarID group. After the map matching is completed, each GPS point has corresponding GIS information, such as road name, road type, city or region where it is located, road direction, etc., to identify the position of the trajectory point on the road. Considering that the intersection approach area to be recognized is usually at a signalized intersection (taking traffic lights as an example in this embodiment), in the matching process of this embodiment, only the vehicle trajectory data in front of the signalized intersection in the road network can be matched or retained, that is, the trajectory data of the waiting vehicles waiting for the red light in front of the traffic lights. Specifically, for the trajectory data of the waiting vehicles at the same signalized intersection, if there are trajectory data corresponding to multiple signalized intersections in the matching information between the vehicle trajectory data and the road network, the waiting vehicle data at each signalized intersection can be obtained respectively, so as to use the trajectory data of the waiting vehicles to identify the approach area at the corresponding intersection.

[0056] Further exemplarily, for the process of obtaining the vehicle trajectory during the red light waiting as described above, the starting moment of red light waiting for each vehicle in the matching information can be recorded (i.e., the starting moment of red light waiting on the road section entering the traffic light intersection (hereinafter referred to as link)), and based on the starting moment of red light waiting and the position of each vehicle, it can be determined whether vehicles in the same import direction but with different turning directions are waiting for the same traffic signal (in a waiting state simultaneously within the same red light cycle), so as to obtain the vehicle trajectory data at the same traffic signal intersection, and save the vehicles that are waiting for the same traffic signal and have trajectory points in both the straight-ahead and left-turn directions, thereby obtaining the corresponding vehicle trajectory data during red light waiting. In this way, the data volume can be effectively reduced and the data accuracy can be improved. In some examples, it is also possible to obtain all the vehicle trajectory data during red light waiting in front of the traffic signal and perform subsequent identification of the merge lane. This embodiment does not make special limitations on this.

[0057] Step S302: Determine the initial merge lane confidence level of the parking position at each time step according to the parking point information at each time step.

[0058] Among them, the initial merge lane confidence level or the merge lane confidence level in the following text is used to represent the credibility that the corresponding parking position is a merge lane. The higher the merge lane confidence level, the higher the reliability that the corresponding parking position is identified as a merge lane. Among them, the initial merge lane confidence level is the confidence level of the parking position at one time step, and the merge lane confidence level is the comprehensive confidence level of the parking position at multiple time steps.

[0059] In this embodiment, by obtaining the parking point information at each time step and using the parking point information at each time step to calculate the initial merge lane confidence level corresponding to the parking position at each time step, and then calculating the comprehensive merge lane confidence level based on the initial merge lane confidence levels at each time step, the identification accuracy of whether the parking position is a merge lane can be effectively improved.

[0060] Exemplarily, the parking point information can carry the area to which the parking position of the vehicle belongs and the number of parking times. The above-mentioned determining the initial merge lane confidence level of the parking position at each time step according to the parking point information at each time step can be implemented in the following manner:

[0061] For each time step, according to the area to which the parking position of the same vehicle among the vehicles at the time step belongs and the number of parking times, determine the initial merge lane confidence level of the parking position at the time step, so as to obtain the initial merge lane confidence level of the parking position at each time step.

[0062] In this embodiment, for any time step, the confidence level of the initial waiting area can be calculated using the area where the parking position is located and the corresponding number of parking times. Exemplarily, taking a time step of 1 hour as an example, for the parking point information obtained from the waiting vehicle trajectory data for each hour, based on the area where the vehicle's parking position is located within that hour and the corresponding number of parking times, the confidence level of the waiting area for the parking position within the corresponding hour is determined. It can be understood that when a vehicle stops waiting for a red light, there are significant differences in the driving characteristics of vehicles turning at traffic lights with a waiting area (such as turning left) and vehicles turning left at traffic lights without a waiting area. Vehicles with a waiting area tend to enter the waiting area part within the intersection when waiting for the light, while vehicles without a waiting area generally stop on the entry link, that is, the straight-ahead section. Therefore, how to determine whether a vehicle is waiting for a red light within the intersection is important for the identification of the waiting area. In this embodiment, it can be determined whether the area where the vehicle's parking position is located includes parking times both in the intersection and on the straight-ahead section, that is, the number of parking times reaches two, to efficiently identify the waiting area. In an alternative approach, for any time step, if the vehicle's parking position (the area it belongs to) only parks on the straight-ahead section or only in the intersection, the initial confidence level of the waiting area can be configured to zero. When the number of vehicles that park both on the straight-ahead section (such as parking once) and in the intersection (such as parking once) reaches a preset ratio (such as 70%, which can be adjusted and determined adaptively according to actual applications) of the total number of vehicles, the initial confidence level of the waiting area is the same as or in a geometric ratio with this preset ratio. For example, if the preset ratio is 70%, the initial confidence level of the waiting area is 0.7. That is, the initial confidence level of the waiting area corresponding to the parking position at the intersection is 0.7.

[0063] In another alternative approach, for any time step, for all vehicles in the waiting vehicle trajectory data, the ratio between the total number of vehicles whose parking position areas include both the straight-ahead section and the intersection and whose number of parking times reaches two (one time on the straight-ahead section and one time in the intersection), and the total number of all vehicles (such as all turning vehicles that have trajectory data both on the straight-ahead section and in the intersection, that is, turning vehicles waiting for the light simultaneously with the straight-ahead vehicles on the straight-ahead section) can be calculated to obtain the confidence level of the waiting area.

[0064] In some alternative approaches, for any time step, other methods can also be used to calculate the initial confidence level of the parking position for the waiting area. For example, by calculating the number of vehicles parking on the straight-ahead section and in the intersection, the higher the number, the higher the confidence level, and so on. It can be understood that the above alternative approaches are only optional examples of this embodiment and do not limit the technical solutions of this application. In other words, this embodiment does not specifically limit the method for calculating the confidence level of the waiting area using the area where the parking position is located and the corresponding number of parking times. Those skilled in the art can make adaptive adjustments in combination with actual applications.

[0065] Next, this embodiment further introduces the above steps of determining the confidence level of the initial waiting area for transfer of the parking position at the time step according to the area to which the parking position of the same vehicle in each vehicle belongs and the number of parking times at the time step, which may include the following steps:

[0066] Obtain a first number of vehicles according to the number of vehicles whose parking positions belong to the straight section and intersection before the traffic light and whose parking times reach twice among the vehicles in the time step;

[0067] According to the number of vehicles having turning trajectory points among the vehicles at the time step and vehicles on the straight section before the traffic light waiting for the light at the same time, a second number of vehicles is obtained;

[0068] The confidence level of the initial waiting-for-turning area of ​​the parking position at the time step is calculated according to the ratio between the first number of vehicles and the second number of vehicles.

[0069] In this embodiment, the area to which the parking position of the vehicle belongs indicates the area range corresponding to the parking position, for example, the area to which the parking position belongs may include a straight section (area) or an intersection (area). For example, in a time step, the area to which the parking position of the vehicle belongs includes both the straight section (area) and the intersection (area) before the traffic light, indicating that the vehicle is a turning vehicle (such as a left-turning vehicle). The turning vehicle stops once in the straight section and once at the intersection, indicating that in addition to stopping at the traffic light, it is also necessary to stop at the intersection. The intersection may have a waiting area for turning. A single vehicle may stop at the intersection due to the vehicle's own factors (such as vehicle failure, operating errors, etc.). Therefore, this embodiment combines the first number of vehicles n at the corresponding time step t. t Calculating the confidence of the initial waiting-to-turn zone can greatly improve the calculation accuracy of the confidence of the waiting-to-turn zone. It should be noted that when a turning vehicle stops at a straight section, it waits for the light at the same time as vehicles on the straight section (such as vehicles without turning trajectory points), that is, left-turning and straight-moving vehicles wait in the same signal cycle. This embodiment uses the characteristic that turning vehicles need to wait for the light at the same time as straight-moving vehicles when passing through traffic lights, and can accurately obtain the total number of turning vehicles, that is, the second vehicle number, that is, mt. For ease of understanding, if Figure 4 As shown in the figure, all vehicles passing through ① (i.e., vehicles waiting for the light at the same time as the straight-moving vehicle) are the second number of vehicles, and all vehicles passing through ② (i.e., vehicles waiting for the light twice at the parking location, which belong to the straight-moving road section and the intersection) are the first number of vehicles. Furthermore, in order to further improve the data progress, outliers can be deleted through clustering according to the stop start time of multiple vehicles to obtain the corresponding first number of vehicles and second number of vehicles.

[0070] Next, by using the first vehicle quantity nt and the second vehicle quantity, the initial confidence level of the stop position in the intersection's initial approach area can be calculated for this time step. In this way, by calculating the number of stops for both the intersection stops and the straight - through section stops, that is, the number of vehicles with two stops, and calculating the initial confidence level of the approach area for this time step by comparing it with the number of vehicles waiting for the light simultaneously for turning and going straight, a more reasonable and accurate calculation method for the confidence level of the approach area can be obtained. In the above - mentioned embodiments, the area to which the stop position belongs can be determined according to the intersection topology information carried in the road network data. For example, if the intersection is marked in the road network, it can be quickly determined whether the area to which the stop position belongs is within the intersection. In some embodiments, considering that the road network data of the SD map obtained in some scenarios may not have the intersection topology information marked, in order to further improve the recognition accuracy of the approach area for unmarked intersections, in this embodiment, the approach area is recognized by fitting trajectory points. Specifically, the method may further include the following steps:

[0071] For each time step, according to the trajectory point data of the same vehicle among the vehicles at this time step, calculate the average speed between any adjacent trajectory points except for the trajectory points corresponding to the straight - through section in front of the traffic light. If the average speed is lower than the preset speed threshold, then obtain the displacement and the moving time difference between the adjacent trajectory points;

[0072] When the displacement is less than the preset distance threshold and the moving time difference between the adjacent trajectory points reaches the preset stay time, determine that the vehicle is in the waiting - for - light state in the stay area corresponding to the adjacent trajectory points in front of the traffic light;

[0073] According to the straight - through section in front of the traffic light and the stay area, obtain the area to which the stop position of the same vehicle among the vehicles at this time step belongs

[0074] For example, during the process of judging the vehicle's second waiting for the light, the vehicle may wait for the light within the intersection. However, if the map does not have the topology information (topo) within the intersection, when performing trajectory - point map matching, it will be difficult to quickly obtain the corresponding number of vehicles using the traditional link - based trajectory matching method. Therefore, in this embodiment, the speed and time information of the trajectory points are used for stop - point recognition (i.e., waiting - for - light behavior recognition), and the Kalman filter method is used to improve the accuracy of trajectory positioning.

[0075] Exemplarily, in this embodiment, the waiting-for-traffic-light behavior is identified through the fitted trajectory points in the trajectory data, and the steps are as follows: Calculate the average speed value of two adjacent trajectory points; when the speed value is lower than the threshold of n kilometers per hour (i.e., the preset speed preset): If it is the first point, record the gps time and the projected position (i.e., the trajectory point position) of the first point. If it is not the first point, determine whether the speed value of the point is lower than the threshold: If it is lower than the threshold, calculate the displacement from the projected position of the first point (i.e., the position between adjacent trajectory points): When the displacement is less than m meters (i.e., the preset distance threshold), it is considered that the vehicle is staying continuously. Calculate the vehicle staying duration (i.e., the moving time difference) by taking the difference between the time of the current gps point and the time of the first point. If the staying duration exceeds t seconds (i.e., the preset staying time), it is considered that the vehicle is waiting and staying in front of the signal light intersection (straight section or intersection), and is in the waiting-for-traffic-light state. If the staying duration does not exceed t seconds, it is not used as a target for calculation. When the displacement is greater than or equal to m meters, it is considered to be slow driving and is not used as a target for calculation. If the speed value is greater than the threshold, calculate the displacement and time difference through the above method. If the displacement is greater than the threshold m: If the staying duration is greater than t seconds, the current moment is the starting moment of the vehicle. If the staying duration is less than or equal to t seconds, it is considered to be an accidental stop and is not used as a target for calculation. For example, in the process of calculating the first vehicle data, the number of vehicles that can reach two stops can be obtained according to the area where the parking position belongs to the straight section in front of the signal light and the staying area.

[0076] Through the above technical solution, the waiting-for-traffic-light state of the vehicle in front of the signal light is judged by the fitting method based on the filtered vehicle trajectory, which can effectively solve the problem that it is difficult to accurately match the gps points at the intersection of the pending turning area in the SD map without the pending turning area topo, and further improve the positioning accuracy of the vehicle at the intersection of the pending turning area.

[0077] Continue to refer to Figure 3 , step S303, determine the pending turning area confidence of the parking position according to the initial pending turning area confidence of the parking position at each time step.

[0078] In this embodiment, the average value between the initial pending turning area confidences of the parking position at each time step can be used to calculate the pending turning area confidence of the parking position. In some embodiments, other methods can also be used to determine the final pending turning area confidence by using the initial pending turning area confidence. For example, weights can be set for the time steps corresponding to different time periods to calculate the final pending turning area confidence, or one or more initial pending turning area confidences corresponding to randomly selected time steps can be used to calculate the final pending turning area confidence. This embodiment does not make special limitations on this.

[0079] In an optional embodiment, the above step S303 determines the confidence level of the waiting area at the parking position according to the initial confidence level of the waiting area at the parking position at each time step, and may adopt the following method: According to the initial confidence level of the waiting area at the parking position at each time step, obtain the distribution of the initial confidence level of the waiting area with respect to the time step through curve fitting to obtain a confidence level distribution curve of the waiting area; when there is no inflection point in the confidence level distribution curve of the waiting area, calculate the first mean value between the initial confidence levels of the waiting area at the parking position at each time step to obtain the confidence level of the waiting area at the parking position.

[0080] In this embodiment, the fitting distribution of the initial confidence level of the waiting area is obtained through curve fitting, and the curve fitting can be performed using a first-order linear function y = ax + b or a second-order linear function y = ax 2 + bx + c or other curve fitting models, and the fitting results are as Figure 5a , Figure 5b and Figure 5c shown, where Figure 5a and Figure 5b are the cases where there is no inflection point in the confidence level distribution curve of the waiting area, and Figure 5c shows the case where there is an inflection point in the confidence level distribution curve of the waiting area.

[0081] For Figure 5a and Figure 5b shown in the case of no inflection point, the calculation formula for the confidence level of the turning area can be as follows:

[0082]

[0083] In the formula, n represents the number of a certain type of vehicle (i.e., the first vehicle number), the condition is waiting for two red lights, and when waiting for the first red light, there are also straight-going vehicles waiting for the light at the same time; m represents the number of vehicles that meet the condition of waiting for the light at the same time as the left-turning vehicle and the straight-going vehicle (i.e., the second vehicle number); t is in hours (i.e., the time step), and it is judged whether there is a waiting area at the intersection according to the confidence level distribution per hour, and the time when the waiting area information starts can be seen from this distribution curve.

[0084] Where the horizontal axis represents a certain time period, and the vertical axis represents the confidence level of the waiting area at the intersection of the entrance direction of a certain traffic light intersection (such as the left-turn direction). The curve is fitted by a multi-segment curve fitting method, and the inflection point of the curve is found. When there is no inflection point in the entire distribution graph, the average confidence level is calculated. When the average confidence level is less than or equal to the k value (such as 0.7, which can be determined according to prior data or empirical values), it is considered that there is no waiting area. When the average confidence level is greater than k, it is considered that there is a waiting area.

[0085] It can be understood that an inflection point is a point where the concavity of a curve changes, that is, a point where the curve changes from concave to convex or from convex to concave. Exemplarily, an inflection point can be determined in the following way. First, calculate the first derivative f(x) of the curve function. The first derivative provides information about the slope of the curve. Then, calculate the second derivative f’(x) of the curve function. The second derivative provides information about the concavity of the curve. An inflection point usually occurs at a point where the second derivative f’(x) is equal to zero. By checking the sign change of the second derivative f’(x) near the point where the second derivative is equal to zero, if the second derivative f’(x) changes from positive to negative, the curve changes from concave to convex, and there is an inflection point. Or if f’(x) changes from negative to positive, the curve changes from convex to concave, and there is an inflection point.

[0086] In some other embodiments, considering that there may be the construction or demolition of the waiting area in practical applications, in order to further improve the freshness of the waiting area, this embodiment can also obtain the change information of the waiting area by using the confidence distribution curve of the waiting area, so as to facilitate the user to identify the changes of the waiting area over a period of time. Specifically, the method provided in this embodiment may further include the following steps: when there is an inflection point in the confidence distribution curve of the waiting area, calculate the second mean value between the initial waiting area confidence levels of the parking position at each time step before the inflection point, and the third mean value between the initial waiting area confidence levels of the parking position at each time step after the inflection point; according to the second mean value and the third mean value, obtain the change information of the waiting area regarding the parking position, and the change information of the waiting area is used to indicate that the waiting area at the intersection changes from existing to non-existing or from non-existing to existing over time.

[0087] As Figure 5c In the scenario with an inflection point as shown, including inflection points a’ and b’ (the time between inflection points a’ and b’ is very short and can be ignored, and inflection points a’ and b are regarded as one inflection point. In some scenarios, there may also be only one inflection point), when calculating the second mean value, the mean value of the initial waiting area confidence levels at each time step before inflection points a’ and b’ can be calculated. Similarly, when calculating the third mean value, the mean value of the initial waiting area confidence levels at each time step after inflection points a’ and b’ can be calculated. The second mean value is around 0.1, and the third mean value is around 0.9. By judging the average confidence level before the inflection point and the average confidence level after the inflection point, when the average confidence level after the inflection point is less than or equal to k, it represents the cancellation of the waiting area, and when it is greater than k, it represents the addition of a new waiting area. In this way, it is possible to quickly determine that the waiting area changes from non-existing to existing.

[0088] It should be noted that the calculation methods of the above second mean value and third mean value can refer to the calculation process of the above first mean value, and relevant descriptions will not be elaborated here.

[0089] Through the above technical solution, users can more intuitively judge the change information of the waiting-turn area, such as whether there is a new waiting-turn area or the demolition of the waiting-turn area. Compared with the static waiting-turn area recognition method, by identifying the overall trend of the confidence level, the dynamic changes of the waiting-turn area can be efficiently judged, thereby improving the freshness of map production.

[0090] Step S304, when the confidence level of the waiting-turn area reaches the preset confidence threshold, determine the intersection waiting-turn area according to the parking position.

[0091] Exemplarily, determining the intersection waiting-turn area according to the parking position that reaches the preset confidence threshold can be to determine the parking position that reaches the preset confidence threshold as the intersection waiting-turn area, or, alternatively, the distribution of the parking positions that reach the preset confidence threshold can also be utilized, such as using the connection lines between the corresponding parking position points to construct the shape topology of the intersection waiting-turn area.

[0092] It should be noted that those skilled in the art can adaptively set the preset confidence threshold in combination with actual applications or prior data, and this embodiment does not make special limitations on the specific threshold.

[0093] The above technical solution provided by this embodiment analyzes the parking point information corresponding to the waiting-vehicle trajectory data at different time steps to obtain the initial confidence level of the waiting-turn area at the parking position at different time steps, thereby determining the confidence level of the waiting-turn area at the parking position. It can achieve the efficient recognition of the intersection waiting-turn area and can adapt to the dynamic changes of the intersection waiting-turn area, thus realizing the efficient recognition of the intersection waiting-turn area. It does not need to rely on manual or collection vehicles to collect the waiting-turn area, effectively solving the problems of low coverage rate and high cost of manual and collection vehicle collection of the waiting-turn area, and facilitating the wide-range extraction of the waiting-turn area for national coverage and all road grades, realizing the recognition of a large number of intersection waiting-turn areas, and having a wider applicability.

[0094] Figure 6 It is a flowchart of another method for identifying an intersection waiting-turn area provided by an embodiment of the present application. On the basis of the above embodiment, this embodiment exemplifies the specific process of obtaining the parking point information. By grouping the waiting-vehicle trajectory data, the parking point information of the straight sections at the signalized intersections and the vehicles within the intersections can be obtained, so as to facilitate distinguishing the regions to which the parking point positions belong, thereby improving the recognition efficiency of the waiting-turn area. In this embodiment, the waiting-vehicle trajectory data is determined based on the matching result between the preset vehicle trajectory data and the road network. The vehicle trajectory data is obtained based on a sliding time window, and the time window is used to indicate a preset time range and moves at this time step. Specifically, in addition to the above steps S301-S304, the method provided by this embodiment may further include the following step S601, and the above step S301 is further divided into step S3011 and step S3012.

[0095] In this embodiment, the vehicle trajectory data is processed by using a sliding time window. For example, the sliding time window can be 1 day and the sliding step can be 1 hour. In some embodiments, the time window and time step can also be adaptively adjusted according to actual applications, and the present application does not make special limitations on this. For each time step, calculate the number of vehicles waiting for the light in front of the intersection and the number of vehicles waiting for the light inside the intersection, and calculate the confidence level of the turn waiting area corresponding to the vehicle stop point information at each time step length accordingly.

[0096] As Figure 6 As shown, step S3011: According to the vehicle trajectory data waiting for the light in front of the signalized intersection, group the vehicle trajectory data waiting for the light according to the straight section and the turning section at the signalized intersection to obtain grouped trajectory data.

[0097] Step S3012: According to the grouped trajectory data, respectively obtain the straight section at the signalized intersection and the stop point information of each vehicle inside the intersection.

[0098] Exemplarily, for the vehicle trajectory data, the trajectory points respectively located in the straight section and the turning section in the trajectory data of the same vehicle can be screened out, so as to obtain the grouped trajectory data respectively located in the straight section and the turning section. And the corresponding stop point information can be obtained through information such as the speed of the trajectory point, which can efficiently determine whether the same vehicle has stop points both in the straight section and inside the intersection. It can be understood that for a vehicle that only has a stop point on the straight section or only inside the intersection, it means that the vehicle only needs to wait for the light once on the straight section and inside the intersection, and there is no need to wait for the light separately inside the intersection (i.e., the turn waiting area at the intersection), indicating that there is no turn waiting area at the current signalized intersection.

[0099] It can be understood that the intersection is the intersection area between the turning section and the straight section, that is, the area where the vehicle leaves the straight section and enters the turning section.

[0100] Through the above technical solution, by grouping the vehicle trajectory data waiting for the light according to the straight section and the turning section, the stop point information on each section can be obtained respectively, so as to facilitate distinguishing the area where the stop point is located, and further improve the efficiency of turn waiting area recognition.

[0101] Furthermore, the above vehicle trajectory data waiting for the light is determined based on the matching result between the preset vehicle trajectory data and the road network, and the following method can be adopted: According to the node information and road segment information in the road network, extract the signalized intersection and the road segment information at the signalized intersection; according to the road segment information at the signalized intersection, filter the trajectory matching data at non-signalized intersections in the matching result to obtain the vehicle trajectory data waiting for the light.

[0102] Optionally, according to the node information and road segment information in the road network, traffic signal intersections and the road segment information at the traffic signal intersections are extracted. The following method can be adopted:

[0103] Exemplarily, the SD map can be compiled to facilitate the extraction of node information and link information in the map road network, and according to this node information and link information, the road segment information at the traffic light intersections is extracted. The road segment information at the traffic light intersections can include intersection ID, entering link ID, exiting link ID, turning information (left turn, right turn, straight, U-turn), link information within the intersection of a complex intersection, and so on. It can be understood that the link information table in the SD map atlas contains the Link information in the map (including information such as link ID, road length, starting node ID, ending node ID, road grade, road width, road shape, etc.), and the node information table contains the point information in the map (including information such as node ID, link ID entering the node, link ID exiting the link, whether it is a traffic light intersection point, Node type: simple intersection, complex intersection, non-intersection point, etc.). The node type can include nodes that are not intersections, that is, ordinary nodes; single intersections, such as Figure 7a the point N shown; the main point of a complex intersection: Figure 7b the hollow point N1 in. The sub-points of a complex intersection: Figure 7b the solid points N2, N3, and N4 in. In this embodiment, the traffic signal intersection can be a single intersection or a complex intersection, and this embodiment does not make a special limitation on this.

[0104] Specifically, in combination with Figure 8 illustrates the process of extracting traffic signal intersections and the road segment information at the traffic signal intersections. By compiling the SD map, node information and link information are obtained. After obtaining the above Node information and Link information, step a can be executed, filtering according to the traffic light attributes at the node nodes, and only retaining the traffic light intersection nodes; and step b can be executed, associating the Node table and the link table, and the association principle is that the entering link and exiting link information of nodid comes from the link table. Then, step c is executed. According to the above association information, for the traffic light intersections, the intersection ID, entering link ID, exiting link ID, turning information (left turn, right turn, straight, U-turn), and link information within the intersection of a complex intersection (such as Figure 7b the links connected between the nodes in) are extracted.

[0105] Through the above steps, the links associated with traffic lights are obtained, and the matching information for the parts of the link that enter, within, and exit the intersection through the traffic light intersection is retained. Since the matching information for the exit link part is not required in the process of identifying the waiting area at the intersection, the matched data can be further filtered according to the traffic light links, and only the trajectory information of the vehicles waiting for the red light before the traffic light intersection is retained. Combining the content described in the above embodiments, the trajectories can be grouped according to the entering link and the turning direction, calculate the starting time of waiting for the light for each vehicle on the entering link, and judge whether the vehicles in different turning directions in the same entrance direction are waiting for the same red light according to the waiting light starting time and position of each vehicle. And only the vehicles waiting for the same light and having vehicle trajectories in both the straight-ahead and left-turn directions can be saved, so as to facilitate obtaining the straight-ahead section at the signal light intersection and the parking points of each vehicle within the intersection. Further combining Figure 9 as shown Figure 9 illustrates the overall process of the waiting area identification process at the intersection of this embodiment, which mainly includes two parts, namely, part A corresponding to real-time trajectory data mining and part B corresponding to offline map data compilation. Among them, real-time trajectory data mining includes steps such as receiving gps point trajectory data, trajectory point matching, traffic light intersection filtering, and waiting area identification, and offline map compilation includes steps such as obtaining map data and extracting traffic light road networks, thereby realizing an efficient waiting area identification process at the intersection.

[0106] In this way, by processing the trajectory data according to the road network nodes and link information, the trajectory data of the vehicles waiting for the light before the signal light intersection is obtained, further improving the accuracy of the trajectory data of the vehicles waiting for the light, providing data support for the identification of the waiting area at the intersection, and thus making the identification of the waiting area more accurate.

[0107] Figure 10 is a schematic structural diagram of a waiting area identification device at an intersection provided by an embodiment of the present application. As Figure 10 shown, the device 1000 includes a first acquisition module 1001, a first determination module 1002, and an identification module 1003, where

[0108] The first acquisition module 1001 is configured to obtain the parking point information of each vehicle at multiple time steps according to the trajectory data of the vehicles waiting for the light before the signal light intersection, and the parking point information is used to indicate the parking position of the vehicle;

[0109] The first determination module 1002 is configured to determine the initial waiting area confidence level at each time step according to the parking point information at each time step;

[0110] An identification module 1003, which is configured to determine the confidence level of the waiting-turn area of the parking position according to the initial confidence level of the waiting-turn area of the parking position at each time step, and when the confidence level of the waiting-turn area reaches a preset confidence level threshold, determine the intersection waiting-turn area according to the parking position.

[0111] In one implementation, the parking point information includes the parking position and the number of parking times of the vehicle; the first determination module 1002 is specifically configured to, for each time step, determine the initial confidence level of the waiting-turn area of the parking position according to the area where the parking position of the same vehicle among the vehicles at the time step belongs and the number of parking times, so as to obtain the initial confidence level of the waiting-turn area of the parking position at each time step.

[0112] In one implementation, the determining the initial confidence level of the waiting-turn area of the parking position according to the area where the parking position of the same vehicle among the vehicles at the time step belongs and the number of parking times includes: obtaining a first vehicle quantity according to the number of vehicles whose parking position area among the vehicles at the time step is the straight section and the intersection in front of the traffic light and the number of parking times reaches two; obtaining a second vehicle quantity according to the number of vehicles with turning trajectory points among the vehicles at the time step and the number of vehicles waiting for the traffic light simultaneously with the vehicles on the straight section in front of the traffic light; calculating the initial confidence level of the waiting-turn area of the parking position at the time step according to the ratio between the first vehicle quantity and the second vehicle quantity.

[0113] In one implementation, the identification module 1003 includes: a curve fitting unit, which is configured to obtain the initial waiting-turn area confidence level distribution with respect to the time step by means of curve fitting according to the initial waiting-turn area confidence level of the parking position at each time step, and obtain a waiting-turn area confidence level distribution curve; a first calculation unit, which is configured to calculate a first mean value between the initial waiting-turn area confidence levels of the parking position at each time step when there is no inflection point in the waiting-turn area confidence level distribution curve, so as to obtain the waiting-turn area confidence level of the parking position.

[0114] In one implementation, the identification module 1003 further includes: a second calculation unit, which is configured to calculate a second mean value between the initial waiting-turn area confidence levels of the parking position at each time step before the inflection point and a third mean value between the initial waiting-turn area confidence levels of the parking position at each time step after the inflection point when there is an inflection point in the waiting-turn area confidence level distribution curve; a change information acquisition unit, which is configured to obtain the waiting-turn area change information about the parking position according to the second mean value and the third mean value, and the waiting-turn area change information is used to indicate that the intersection waiting-turn area changes from existing to non-existing or from non-existing to existing over time.

[0115] In one embodiment, the device further includes: a second acquisition module configured to calculate, for each time step, an average speed between any adjacent trajectory points except for the trajectory points corresponding to the straight section before the traffic light based on the trajectory point data of the same vehicle in each vehicle. If the average speed is lower than a preset speed threshold, the displacement and the moving time difference between the adjacent trajectory points are acquired; a second determination module configured to determine that the vehicle is in a waiting-for-traffic-light state in the waiting area corresponding to the adjacent trajectory points when the displacement is less than a preset distance threshold and the moving time difference between the adjacent trajectory points reaches a preset waiting time; a third acquisition module configured to acquire the area to which the parking position of the same vehicle in each vehicle belongs at the time step based on the straight section before the traffic light and the waiting area.

[0116] In one embodiment, the waiting-for-traffic-light vehicle trajectory data is determined based on a matching result between pre-set vehicle trajectory data and a road network, wherein the vehicle trajectory data is obtained based on a sliding time window, and the time window is used to indicate a preset time range and moves in time steps; the first acquisition module 1001 includes: a grouping unit configured to group the waiting-for-traffic-light vehicle trajectory data according to the straight section and the turning section at the traffic light intersection to obtain grouped trajectory data; a parking point acquisition unit configured to respectively acquire the straight section at the traffic light intersection and the parking point information of each vehicle in the intersection according to the grouped trajectory data.

[0117] In one embodiment, the first acquisition module 1001 is further configured to determine the waiting-for-traffic-light vehicle trajectory data based on a matching result between pre-set vehicle trajectory data and a road network. The first acquisition module further includes: an extraction unit configured to extract the traffic light intersection and the road section information at the traffic light intersection according to the node information and the road section information in the road network; a matching unit configured to filter the trajectory matching data at non-traffic-light intersections in the matching result according to the road section information at the traffic light intersection to obtain the waiting-for-traffic-light vehicle trajectory data.

[0118] For relevant descriptions and effects, reference may be made to the corresponding descriptions of the steps in the method embodiments of the present application, and no further elaboration will be made here.

[0119] Figure 11 The figure is a schematic structural diagram of an electronic device provided in an embodiment of the present application. As Figure 10 shown, the electronic device includes: a processor 1101 and a memory 1102 communicatively connected to the processor 1101;

[0120] The memory 1102 stores computer-executable instructions;

[0121] The processor 1101 executes the computer-executable instructions stored in the memory 1102 to implement the intersection left-turn waiting area recognition method. Among them, the memory 1102 and the processor 1101 are connected through a bus 1103.

[0122] For relevant descriptions, reference can be made to the corresponding descriptions and effects of the steps in the method embodiments of this application for understanding, and details are not elaborated here.

[0123] The embodiments of this application further provide a computer-readable storage medium storing computer-executable instructions, which are used to implement the intersection left-turn waiting area recognition method provided in the above method embodiments when executed by a processor.

[0124] Among them, the computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0125] The embodiments of this application further provide a computer program product, which includes a computer program that implements the intersection left-turn waiting area recognition method provided in the above method embodiments when executed by a processor.

[0126] For relevant descriptions, reference can be made to the corresponding descriptions and effects of the steps in the method embodiments of this application for understanding, and details are not elaborated here.

[0127] The embodiments of this application further provide a chip, including a memory and a processor. The memory is used to store a computer program, and the processor is used to call and run the computer program from the memory to execute the intersection left-turn waiting area recognition method.

[0128] For relevant descriptions, reference can be made to the corresponding descriptions and effects of the steps in the method embodiments of this application for understanding, and details are not elaborated here.

[0129] In several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of modules is only a logical function division. In actual implementation, there can be other division methods. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of devices or modules can be in electrical, mechanical or other forms.

[0130] Other embodiments of the present application will be readily apparent to those skilled in the art upon consideration of the specification and practice of the application disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include known common general knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and examples are only to be considered as exemplary, and the true scope and spirit of the present application are pointed out by the following claims.

[0131] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present application is only limited by the appended claims.

Claims

1. A method for identifying a left-turn waiting area at an intersection, characterized in that, Including: Obtain the stop point information of each vehicle at multiple time steps according to the trajectory data of the waiting vehicles in front of the signalized intersection, where the stop point information is used to indicate the stop position of the vehicle. Determine the initial confidence level of the stop position in the approach lane at each time step according to the stop point information at each time step. Determine the confidence level of the stop position in the approach lane according to the initial confidence level of the stop position in the approach lane at each time step, and when the confidence level of the approach lane reaches the preset confidence threshold, determine the approach lane of the intersection according to the stop position.

2. The method according to claim 1, wherein The stop point information includes the stop position and the number of stops of the vehicle. The determining the initial confidence level of the stop position in the approach lane at each time step according to the stop point information at each time step includes: For each time step, determine the initial confidence level of the stop position in the approach lane at the time step according to the area where the stop position of the same vehicle among the vehicles at the time step belongs and the number of stops, so as to obtain the initial confidence level of the stop position in the approach lane at each time step.

3. The method according to claim 2, characterized in that, The determining the initial confidence level of the stop position in the approach lane at the time step according to the area where the stop position of the same vehicle among the vehicles at the time step belongs and the number of stops includes: Obtain the first vehicle number according to the number of vehicles whose stop position area among the vehicles at the time step is the straight section and the intersection in front of the signal lamp and the number of stops reaches two. Obtain the second vehicle number according to the number of vehicles with turning trajectory points among the vehicles at the time step and the number of vehicles waiting for the signal at the same time on the straight section in front of the signal lamp. Calculate the initial confidence level of the stop position in the approach lane at the time step according to the ratio between the first vehicle number and the second vehicle number.

4. The method according to any one of claims 1 to 3, characterized in that, The determining the confidence level of the stop position in the approach lane according to the initial confidence level of the stop position in the approach lane at each time step includes: Obtain the distribution of the initial confidence level of the approach lane with respect to the time step by curve fitting according to the initial confidence level of the stop position in the approach lane at each time step, and obtain the confidence level distribution curve of the approach lane. When there is no inflection point in the confidence level distribution curve of the approach lane, calculate the first mean value between the initial confidence levels of the stop position in the approach lane at each time step, and obtain the confidence level of the stop position in the approach lane.

5. The method according to claim 4, wherein It also includes: When there is an inflection point in the confidence level distribution curve of the approach lane, calculate the second mean value between the initial confidence levels of the stop position in the approach lane at each time step before the inflection point, and the third mean value between the initial confidence levels of the stop position in the approach lane at each time step after the inflection point. Obtain the change information of the approach lane with respect to the stop position according to the second mean value and the third mean value, where the change information of the approach lane is used to indicate that the approach lane of the intersection changes from existing to non-existing or from non-existing to existing over time.

6. The method according to any one of claims 1-5, characterized in that, It also includes: For each time step, according to the trajectory point data of the same vehicle in each vehicle at the time step, the average speed between any adjacent trajectory points except the trajectory point corresponding to the straight section before the traffic light is calculated, and if the average speed is lower than the preset speed threshold, the displacement and moving time difference between the adjacent trajectory points are obtained; When the displacement is less than a preset distance threshold and the moving time difference between the adjacent track points reaches a preset stay time, it is determined that the vehicle is in a waiting state before a traffic light in the stay area corresponding to the adjacent track points; According to the straight road section before the traffic light and the stop area, the area to which the parking position of the same vehicle among the vehicles in the time step belongs is obtained.

7. The method according to any one of claims 1-5, characterized in that, The vehicle trajectory data waiting for the light is determined based on the matching result between the preset vehicle trajectory data and the road network, wherein the vehicle trajectory data is obtained based on a sliding time window, the time window is used to indicate a preset time range and moves at the time step; The method of obtaining the parking point information of each vehicle at multiple time steps based on the trajectory data of vehicles waiting for the light at the traffic light intersection includes: According to the trajectory data of vehicles waiting for the light before the traffic light intersection, the trajectory data of vehicles waiting for the light are grouped according to the straight section and the turning section at the traffic light intersection to obtain grouped trajectory data; According to the grouped trajectory data, the straight section at the traffic light intersection and the parking point information of each vehicle in the intersection are respectively obtained.

8. The method according to claim 7, wherein The vehicle trajectory data waiting for the light is determined based on the matching result between the preset vehicle trajectory data and the road network, including: Extracting the signal light intersection and the road segment information at the signal light intersection according to the node information and the road segment information in the road network; According to the road segment information at the signal light intersection, the trajectory matching data at the non-signal light intersection in the matching results are filtered to obtain the trajectory data of the vehicle waiting for the light.

9. An intersection waiting area recognition device, characterized in that, include: A first acquisition module is configured to acquire parking point information of each vehicle at multiple time steps based on trajectory data of vehicles waiting for the light at a traffic light intersection, wherein the parking point information is used to indicate a parking position of the vehicle; A first determination module is configured to determine the initial waiting area confidence of the parking position at each time step according to the parking point information at each time step; An identification module is configured to determine the confidence of the waiting area for the parking position according to the initial confidence of the waiting area for the parking position at each time step, and when the confidence of the waiting area for the parking position reaches a preset confidence threshold, determine the waiting area for the intersection according to the parking position.

10. An electronic device / computer-readable storage medium / computer program product, characterized in that include: Memory and processor; The memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory, so that the electronic device executes the method for identifying a turn-around area at an intersection as described in any one of claims 1 to 9; and / or The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method for identifying a turn-around area at an intersection as described in any one of claims 1 to 8; and / or The computer program product includes a computer program which, when executed by a processor, implements the intersection left-turn waiting area recognition method according to any one of claims 1-8.

Images