Trajectory data processing method and device, equipment and storage medium

Abstract

This disclosure provides a trajectory data processing method, apparatus, device, and storage medium, relating to the field of data processing, particularly to the fields of artificial intelligence, big data, autonomous driving, and mapping. The specific implementation scheme is as follows: Based on multiple driving trajectories of a vehicle, an intersection segment in the target road is determined; based on the time interval between trajectory points corresponding to the intersection segment, the target stopping duration corresponding to the intersection segment is determined; if the target stopping duration corresponding to the intersection segment is greater than a preset stopping duration, effective stopping points and the target number of effective stopping points are determined from multiple trajectory points; based on the target number of effective stopping points, initial prediction information for the intersection segment is obtained, wherein the initial prediction information is used to indicate the presence of traffic signal equipment; target image data corresponding to the intersection segment is determined; based on the target image data, the initial prediction information is verified to obtain the target verification result.

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, and in particular to the fields of artificial intelligence, big data, autonomous driving, and mapping. Background Technology

[0002] With the deep integration of intelligent transportation systems and high-precision maps, data from traffic signal devices (such as traffic lights, or simply traffic signal data) is a key basis for navigation route planning, voice broadcasting, and intelligent driving of vehicles. The accuracy of traffic signal data has a crucial impact on the application of map data.

[0003] Currently, the traffic light data recognition and update process still has significant shortcomings. For example, in cloud-based image recognition traffic light solutions, there are problems such as large end-to-end latency (affected by network transmission, etc.), poor timeliness, high image storage and transmission costs, high recognition resource occupancy, and inability to dynamically determine the status of traffic signal equipment (for example, in scenarios where there are traffic lights but they are not lit, dynamic determination is not possible). In vehicle-side image recognition traffic light scenarios, there are problems such as limited coverage of high-timeliness update road sections, low recognition accuracy in complex roads and environments (for example, vehicle-side recognition models require lightweight design, thus limiting performance and resulting in low recognition accuracy), and inability to dynamically determine the status of traffic signal equipment. Summary of the Invention

[0004] This disclosure provides a trajectory data processing method, apparatus, device, and storage medium.

[0005] According to one aspect of this disclosure, a trajectory data processing method is provided, comprising: Based on multiple driving trajectories of the vehicle, the intersections and road segments in the target road are identified; each driving trajectory contains multiple trajectory points. Based on the time interval between the trajectory points corresponding to the intersection and road segment, the target dwell time corresponding to the intersection and road segment is determined; If the target dwell time corresponding to the intersection segment is greater than the preset dwell time, determine the effective dwell points and the target number of effective dwell points from multiple trajectory points; Based on the target number of effective stopping points, initial prediction information is obtained for the intersection segment, wherein the initial prediction information is used to indicate whether traffic signal equipment exists; The target image data corresponding to the intersection segment is determined, and the initial prediction information is verified based on the target image data to obtain the target verification result; wherein, the target verification result is used to characterize the accuracy of the initial prediction information.

[0006] According to another aspect of this disclosure, a trajectory data processing apparatus is provided, comprising: The trajectory data processing unit is used to determine the intersection segment in the target road based on multiple driving trajectories of the vehicle; wherein each driving trajectory contains multiple trajectory points; based on the time interval between the trajectory points corresponding to the intersection segment, determine the target stopping duration corresponding to the intersection segment; if the target stopping duration corresponding to the intersection segment is greater than a preset stopping duration, determine the effective stopping points and the target number of effective stopping points from the multiple trajectory points; based on the target number of effective stopping points, obtain initial prediction information for the intersection segment, wherein the initial prediction information is used to indicate whether traffic signal equipment exists; The prediction result verification unit is used to determine the target image data corresponding to the intersection segment, and to verify the initial prediction information based on the target image data to obtain the target verification result; wherein, the target verification result is used to characterize the accuracy of the initial prediction information.

[0007] According to another aspect of this disclosure, an electronic device is provided, comprising: At least one processor; and The memory is communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform any of the methods described in the present disclosure.

[0008] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause the computer to perform any of the methods according to embodiments of this disclosure.

[0009] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements any of the methods according to embodiments of this disclosure.

[0010] This disclosure provides a method for identifying and verifying traffic signal equipment by integrating vehicle trajectory and image data, aiming to overcome the limitations of traditional methods in terms of update timeliness, coverage, and dynamic perception. Specifically, this disclosure can determine the intersection segment of the target road based on the vehicle's driving trajectory, determine the target stopping time corresponding to the intersection segment, and then determine the effective stopping points and the target number of effective stopping points to obtain initial prediction information for the intersection segment. Finally, the initial prediction information is verified by combining the target image data corresponding to the intersection segment. Through joint discrimination of trajectory stopping features and target image data, accurate discovery and reliable verification of traffic signal equipment at intersection segments are achieved, improving the recall rate and accuracy of traffic signal equipment identification.

[0011] Furthermore, because this disclosed solution can utilize widely available trajectory data for preliminary mining, it achieves low-cost, wide-area, and high-frequency discovery of traffic signal devices, overcoming the cost and coverage bottlenecks of existing purely visual surveys. It also significantly shortens the data update cycle and improves timeliness. Moreover, because this disclosed solution can use driving trajectories for initial mining to obtain a candidate set with high recall, and then combine this with visual technology for verification to ensure a high-precision final result, the above combination effectively smooths out performance and cost, thereby solving the problems of low timeliness, high cost, and limited coverage of traditional methods.

[0012] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0013] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein: Figure 1 This is an illustrative flowchart of a trajectory data processing method according to an embodiment of this application. Figure 1 ; Figure 2 This is an illustrative flowchart of a trajectory data processing method according to an embodiment of this application. Figure 2 ; Figure 3 This is an illustrative schematic diagram of trajectory data according to an embodiment of this application; Figure 4 This is an illustrative flowchart of a trajectory data processing method according to an embodiment of this application. Figure 3 ; Figure 5 This is a schematic diagram of the result of trajectory data processing according to an embodiment of this application; Figure 6 This is a schematic diagram of the implementation flow of a trajectory data processing method according to an embodiment of this application in a specific example; Figure 7 This is a schematic diagram of the structure of a trajectory data processing device according to an embodiment of this application; Figure 8 This is a block diagram of an electronic device used to implement the trajectory data processing method of the embodiments of this disclosure. Detailed Implementation

[0014] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0015] In this document, the term "and / or" merely describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. The term "at least one" in this document indicates any combination of at least two of a plurality of elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C. The terms "first" and "second" in this document refer to and distinguish between multiple similar technical terms, not to restrict the order or to limit there to only two. For example, "first feature" and "second feature" refer to two categories / two features; the first feature can be one or more, and the second feature can also be one or more.

[0016] Furthermore, to better illustrate this disclosure, numerous specific details are set forth in the following detailed description. Those skilled in the art will understand that this disclosure can still be practiced even without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art have not been described in detail in order to highlight the main points of this disclosure.

[0017] The following describes the related technologies of the embodiments of this disclosure. The following related technologies are optional solutions and can be combined with the technical solutions of the embodiments of this disclosure in any way, and they all fall within the protection scope of the embodiments of this disclosure.

[0018] Traffic lights (also known as signal lights) correspond to traffic lights in the real world. In map data, they exist as a separate data layer, primarily used for optimal route selection in navigation route planning, determining whether routes pass through fewer traffic lights, and calculating route travel time. They are also used for voice prompts, such as reminding drivers to "turn right at the second traffic light ahead." In autonomous driving, they assist vehicles in intersection positioning, starting and stopping at intersections, and turning at intersections based on light status. Therefore, the accuracy of data information such as the presence or absence of traffic lights, the intersections they affect, the direction of the controlled road segments, and their lighting status is crucial for navigation route planning, voice prompts, and autonomous driving.

[0019] Currently, the traffic light data recognition and update process still has significant shortcomings. For example, in cloud-based image recognition traffic light solutions, there are problems such as large end-to-end latency (affected by network transmission, etc.), poor timeliness, high image storage and transmission costs, high recognition resource occupancy, and inability to dynamically determine the status of traffic signal equipment (for example, in scenarios where there are traffic lights but they are not lit, dynamic determination is not possible). In vehicle-side image recognition traffic light scenarios, there are problems such as limited coverage of high-timeliness update road sections, low recognition accuracy in complex roads and environments (for example, vehicle-side recognition models require lightweight design, thus limiting performance and resulting in low recognition accuracy), and inability to dynamically determine the status of traffic signal equipment.

[0020] Based on this, the present application provides a trajectory data processing method that can determine the trajectory stopping characteristics of intersection segments based on vehicle trajectory data. Then, based on the trajectory stopping characteristics and combined with the joint discrimination of target image data, it can achieve accurate mining and reliable verification of traffic signal equipment at intersection segments, effectively improving the recall rate and accuracy of traffic signal equipment identification. At the same time, it is timely and has low image acquisition cost.

[0021] Specifically, Figure 1 This is an illustrative flowchart of a trajectory data processing method according to an embodiment of this application. Figure 1 This method can be optionally applied to electronic devices, such as personal computers, servers, server clusters, and other electronic devices.

[0022] Furthermore, the method includes at least a portion of the following: For example... Figure 1 As shown, it includes: Step S101: Based on multiple driving trajectories of the vehicle, determine the intersections and road segments in the target road.

[0023] Here, each driving trajectory contains multiple trajectory points.

[0024] It should be noted that the multiple driving trajectories in this disclosed solution can be specifically multiple driving trajectories of a single vehicle at different times, or they can be specifically driving trajectories of multiple vehicles at one time or multiple different time periods. In other words, this disclosed solution does not impose specific restrictions on driving trajectories.

[0025] Furthermore, in one example, the driving trajectory can specifically be continuous driving data collected through vehicle positioning modules, roadside sensing devices, navigation platforms, etc. Furthermore, in one example, each driving trajectory contains multiple discretely distributed trajectory points.

[0026] Furthermore, in one example, the trajectory point may include one or more of the following: vehicle location information, acquisition timestamp, vehicle motion state (such as driving speed, driving heading angle, etc.), and contextual information (such as device identifier, trajectory point serial number, quantity source, etc.). Further, the vehicle location information may specifically be the coordinate information of the trajectory point in a specified coordinate system (e.g., the standard coordinate system corresponding to the positioning system, the world coordinate system, etc.), thus effectively ensuring the accuracy and uniqueness of the trajectory point's spatial location.

[0027] Furthermore, in another example, the target road is a road network with vehicle traffic function, such as urban municipal roads, intercity highways, rural roads, and internal roads of industrial parks. This disclosure does not impose specific restrictions on the target road.

[0028] Furthermore, in one example, the intersection segment can specifically refer to the passage area where the target road intersects with other roads and may have traffic signal equipment, such as a crossroads, T-junctions, roundabouts, and other road segments with intersection characteristics.

[0029] Furthermore, in one example, the spatial range of an intersection segment can be defined by a preset distance threshold (e.g., the area from the center point of the intersection to a distance of 50m from that center point), thereby facilitating accurate matching of trajectory points with the intersection segment. For example, trajectory points can be matched with intersection segments in the following way: based on the preset distance threshold corresponding to the intersection segment, the spatial influence range of the intersection segment is determined, and trajectory points falling within the spatial influence range of the intersection segment are determined as trajectory points matched with the intersection segment, that is, determined as trajectory points corresponding to the intersection segment.

[0030] Step S102: Determine the target dwell time corresponding to the intersection segment based on the time interval between the trajectory points corresponding to the intersection segment.

[0031] Here, in one example, the time interval between the trajectory points mentioned above can be specifically the time interval between two adjacent trajectory points in the same vehicle's driving trajectory within the intersection segment, such as the difference in the collection timestamps corresponding to two adjacent trajectory points. This makes it easier to statistically determine the continuous dwell time of a single vehicle within the intersection segment.

[0032] It should be noted that, in practical applications, in order to improve the accuracy of the target's dwell time and enable it to effectively express the dwelling behavior, thereby improving the accuracy of the recognition results, this disclosed solution can also filter vehicles. For example, it can eliminate trajectory interference from trucks, pedestrians, etc., and select the driving trajectory of vehicles whose vehicle type meets the preset requirements as the driving trajectory of this application solution. For example, it can select the driving trajectory of a small car as the driving trajectory required by this application solution in the future.

[0033] Furthermore, in order to further improve the accuracy of the target dwell time in expressing the dwelling behavior and avoid the bias caused by sparse sampling, the present application can further filter the driving trajectory to select the driving trajectory with a trajectory point density greater than the preset density requirement, so as to use the driving trajectory required by the present application in the subsequent process. In this way, the problem of inaccurate prediction results caused by sparse sampling can be avoided.

[0034] Furthermore, in order to further ensure that more trajectory points in the driving trajectory fall within the intersection segment to be identified by the traffic signal equipment, the coverage of trajectory points in the intersection segment can be further determined, and then the driving trajectory with a trajectory point coverage higher than the preset coverage can be selected as the driving trajectory required for subsequent use in this application scheme.

[0035] Furthermore, in one example, the target dwell time mentioned above can specifically refer to the historical dwell time. For example, in one example, the target dwell time is obtained based on the following information: the time interval between two adjacent trajectory points in the intersection segment based on the driving trajectories of different vehicles (e.g., the driving trajectories of different vehicles at the same time period, or the driving trajectories of different vehicles at different time periods) or the driving trajectories of the same vehicle at different time periods. This is used to represent the historical dwell characteristics of different vehicles, or the same vehicle at different time periods, in the intersection segment.

[0036] It should be noted that in practical applications, the statistical method for the target dwell time can be determined based on the specific type of intersection and road segment, traffic conditions, etc., so as to provide stable and quantifiable data support for subsequent identification of traffic signal equipment based on the target dwell time.

[0037] Step S103: If the target stopping time corresponding to the intersection segment is greater than the preset stopping time, determine the effective stopping points and the number of targets with effective stopping points from multiple trajectory points.

[0038] It should be noted that the multiple trajectory points used to filter valid stopping points in step S103 can specifically be the trajectory points used to calculate the target stopping time. For example, in one example, for a determined driving trajectory, if every trajectory point in the driving trajectory participates in the calculation of the target stopping time, then every trajectory point in the driving trajectory is used as a trajectory point for filtering in step S103. Otherwise, if there are trajectory points in the driving trajectory that do not participate in the calculation of the target stopping time, then these non-participating trajectory points are removed from the total number of trajectory points. This provides favorable support for improving the accuracy of valid stopping point identification and the accuracy of the number of valid stopping points.

[0039] It should be noted that the preset dwell time is an empirical value, which can be determined based on the historical dwelling behavior at known traffic light intersections in actual scenarios. This public solution does not impose specific restrictions on it.

[0040] Step S104: Based on the target number of effective stopping points, obtain initial prediction information for the intersection segment.

[0041] Here, the initial prediction information is used to indicate whether traffic signal equipment is present.

[0042] In other words, in this example, it is possible to predict whether there are traffic signal devices at the intersection based on the number of target valid stopping points.

[0043] Step S105: Determine the target image data corresponding to the intersection segment, and verify the initial prediction information based on the target image data to obtain a target verification result. Here, the target verification result is used to characterize the accuracy of the initial prediction information.

[0044] In one example, the target image data can specifically be images collected by a vehicle while it is traveling at the intersection, or images or video frames collected by roadside monitoring equipment or inspection vehicles within the intersection.

[0045] It should be noted that the target image data must match the trajectory data used in the scheme. For example, the two should be spatiotemporally aligned, meaning that the acquisition time and location of the target image data should match the timestamps and spatial locations of the trajectory points contained in the driving trajectory used in the scheme. This avoids problems such as inaccurate verification results due to spatiotemporal inconsistencies.

[0046] Furthermore, in one example, the target verification result can further characterize the state of the traffic signal equipment (e.g., real-time status), such as the on / off status of the traffic lights and their current color. In other words, this disclosed solution, through the synergy of "trajectory feature inference + real-time image verification," can both discover the location of the equipment and judge its dynamic working status, effectively improving the dynamic perception capability of traffic signal equipment, and thus providing favorable support for the subsequent promotion of intelligent transportation and smart cities.

[0047] Thus, this disclosed solution provides a method for identifying and verifying traffic signal equipment by integrating vehicle trajectory and image data, aiming to overcome the limitations of traditional solutions in terms of update timeliness, coverage, and dynamic perception. Specifically, this disclosed solution can determine the intersection segment of the target road based on the vehicle's driving trajectory, determine the target stopping duration corresponding to the intersection segment, and then determine the effective stopping points and the target number of effective stopping points to obtain initial prediction information for the intersection segment. Finally, the initial prediction information is verified by combining the target image data corresponding to the intersection segment. Through joint discrimination of trajectory stopping features and target image data, accurate discovery and reliable verification of traffic signal equipment at intersection segments are achieved, improving the recall rate and accuracy of traffic signal equipment identification.

[0048] Furthermore, because this disclosed solution can utilize widely available trajectory data for preliminary mining, it achieves low-cost, wide-area, and high-frequency discovery of traffic signal devices, overcoming the cost and coverage bottlenecks of existing purely visual surveys. It also significantly shortens the data update cycle and improves timeliness. Moreover, because this disclosed solution can use driving trajectories for initial mining to obtain a candidate set with high recall, and then combine this with visual technology for verification to ensure a high-precision final result, the above combination effectively smooths out performance and cost, thereby solving the problems of low timeliness, high cost, and limited coverage of traditional methods.

[0049] Furthermore, in a specific example, the intersection segment in the target road can be determined in the following way; specifically, the determination of the intersection segment in the target road based on multiple driving trajectories of the vehicle (e.g., step S101) can specifically include: Step S101-1: Match the multiple driving trajectories of the vehicle with the preset road network data to determine the actual road to which each driving trajectory data belongs.

[0050] Here, the preset road network data refers to pre-determined, standardized road vector data, such as map road network data for a specified area. It should be noted that this disclosure does not limit the specific preset road network data.

[0051] Furthermore, after determining the preset road network data, matching can be performed in the following ways, such as, but not limited to, one or more of spatial projection matching, topological path matching, and spatial nearest neighbor matching. In this way, the positioning information of the trajectory points is mapped to the preset road network data, thereby accurately locating the actual road corresponding to each driving trajectory, which provides favorable support for the subsequent accurate discovery of traffic signal equipment at intersections and road sections.

[0052] Step S101-2: Based on the number of trajectory points in the actual road, filter out the actual roads with a trajectory coverage rate higher than the preset coverage rate threshold as target roads.

[0053] Here, the preset coverage threshold can be set according to the actual application situation, and this disclosure does not impose specific restrictions on it.

[0054] Furthermore, it should be noted that the trajectory coverage rate is an indicator that quantifies the degree to which the driving trajectory covers the actual road. For example, it can be specifically represented by the total number of trajectory points contained within a unit length of the actual road. In this case, the higher the trajectory coverage rate, the more trajectory points there are on the actual road, which makes it easier to accurately mine the data. This can effectively avoid the problem of inaccurate feature judgment caused by a small number of trajectories, and provide data support for the subsequent accurate mining of traffic signal equipment at intersections and road sections and the improvement of the accuracy of prediction results.

[0055] For example, in one instance, different trajectory statistics strategies can be employed for road segments with varying traffic volumes to avoid biases caused by insufficient trajectory point data. For instance, for high-traffic road segments, the trajectory coverage rate can be determined directly using the daily trajectory coverage; while for low-traffic road segments, the cumulative trajectory coverage over multiple days can be used. In other words, this disclosed solution employs differentiated statistical periods (single-day or multi-day cumulative) for road segments with high and low traffic volumes, effectively solving the problem of insufficient daily data for low-traffic road segments and enhancing the robustness and adaptability of this disclosed solution in different scenarios. This effectively prevents distortions in coverage assessment and improper target road selection due to insufficient trajectory samples.

[0056] Step S101-3: Determine the intersection segment based on the intersection location in the target road.

[0057] Here, the road length of the intersection segment is shorter than the road length of the target road. In other words, this disclosed solution extracts a shorter "intersection segment" from the "target road" based on the "intersection location". This step narrows the analysis scope from the entire road to the intersection area where traffic behavior is most complex and requires the most attention, making subsequent high-precision mining and analysis of traffic signal equipment more efficient.

[0058] This proposed solution introduces trajectory coverage as a core screening indicator to select target roads. This automatically filters out roads with sparse data and lacking statistical significance, allowing subsequent processing logic to directly focus on "target roads" with high traffic volume and abundant data, significantly improving computational efficiency and result reliability. Furthermore, after identifying the target intersection, this solution further delineates shorter intersection segments based on the intersection's location. This narrows the analysis scope from the entire road to the intersection area with the most complex traffic behavior and the greatest need for attention, effectively reducing computational load and making subsequent high-precision analysis of traffic signal equipment more efficient.

[0059] Figure 2 This is an illustrative flowchart of a trajectory data processing method according to an embodiment of this application. Figure 2 This method can be optionally applied to electronic devices, such as personal computers, servers, and server clusters. It is understood that the above... Figure 1 The methods shown can also be applied to this example, and the related content will not be elaborated further in this example.

[0060] Furthermore, the method includes at least a portion of the following: For example... Figure 2 As shown, it includes: Step S201: Based on multiple driving trajectories of the vehicle, determine the intersections and road segments in the target road.

[0061] Here, each driving trajectory contains multiple trajectory points.

[0062] Step S202: Based on the time interval between the trajectory points corresponding to the intersection and road segment, determine whether there is any stopping behavior of the trajectory points.

[0063] For example, in one example, step S202 can be specifically described as follows: for multiple trajectory points of the same driving trajectory of the same vehicle falling into the intersection section, determine whether the trajectory point has stopped behavior based on the time interval between two adjacent trajectory points.

[0064] Specifically, in one example, if the time interval between two adjacent trajectory points is greater than a preset duration (e.g., 30 seconds), it can be considered that the two adjacent trajectory points exhibit stopping behavior. Otherwise, if the time interval between two adjacent trajectory points is less than or equal to the preset duration (e.g., 30 seconds), it can be considered that neither of the two adjacent trajectory points exhibits stopping behavior.

[0065] Alternatively, in another example, to further improve the accuracy of judging the stopping behavior, the actual distance between two adjacent trajectory points can be used as the judgment criterion; in this case, step S202 described above can specifically include: Based on the time interval between two adjacent trajectory points of the same driving trajectory corresponding to the intersection and road segment, and the actual distance between the two adjacent trajectory points, it is determined whether the trajectory point has stopped.

[0066] For example, in one example, if the time interval between two adjacent trajectory points of the same driving trajectory corresponding to the intersection segment is greater than a preset duration, and the actual distance between the two adjacent trajectory points is less than a preset length (e.g., within 10 meters), then it can be considered that the two adjacent trajectory points have stopped.

[0067] For example, in one example, such as Figure 3 As shown, for the same vehicle's driving trajectory T1, the driving trajectory T1 includes the following: Figure 3 The three trajectory points shown are, for example, trajectory point 1, trajectory point 2, and trajectory point 3. If the actual distance between trajectory point 1 and trajectory point 2 is less than or equal to 10 meters, and the time interval between trajectory point 1 and trajectory point 2 exceeds 30 seconds, then trajectory point 1 and trajectory point 2 are considered to have stopped.

[0068] Step S203: The trajectory points that exhibit stopping behavior within a preset time window are taken as stopping points to obtain multiple stopping points corresponding to the intersection segment.

[0069] Here, the preset time window can specifically refer to a preset time interval. For example, it can be flexibly configured according to factors such as actual road conditions and data collection frequency to filter trajectory points with stopping behavior.

[0070] For example, in a specific example, the traffic signal device could be a traffic light. In this case, the preset time window can be determined based on factors such as the green light cycle and the red light cycle. For instance, in a scenario where the green light cycle is 40-80 seconds and the red light cycle is 60-120 seconds, the preset time window could be set to 3 minutes. Then, the trajectory points exhibiting stopping behavior within a preset time window, such as 3 minutes, can be used as stopping points. For example, continuing with... Figure 3 For example, for driving trajectory T1, within a preset time window, driving trajectory T1 includes the following: Figure 3 The three trajectory points shown are trajectory point 1, trajectory point 2, and trajectory point 3. Correspondingly, the time interval between adjacent trajectory points 1 and 2 is denoted as B1, and the time interval between adjacent trajectory points 2 and 3 is denoted as B2. Further, if the time interval B1 between trajectory points 1 and 2 satisfies the above condition, trajectory points 1 and 2 are designated as stopping points, for example, denoted as stopping point 1 and stopping point 2. Similarly, if the time interval B2 between trajectory points 2 and 3 satisfies the above condition, trajectory point 3 is also designated as a stopping point, for example, denoted as stopping point 3.

[0071] It should be noted that in practical applications, in order to further increase the number of stopping points and thus increase the information density that the target stopping time can carry, the trajectory points of stopping behavior within multiple preset time windows can be used as stopping points, thereby obtaining the stopping points within multiple preset time windows corresponding to the intersection segment.

[0072] Step S204: Determine the target stopping time corresponding to the intersection segment based on the time interval between stopping points.

[0073] It should be noted that the time interval between stopping points can be specifically the time interval between two adjacent stopping points within the intersection segment of the same vehicle's driving trajectory. For example, the difference in the collection timestamps corresponding to two adjacent stopping points. This makes it easier to statistically determine the target stopping duration corresponding to the intersection segment.

[0074] Step S205: If the target stopping time corresponding to the intersection segment is greater than the preset stopping time, determine the effective stopping points and the number of targets with effective stopping points from multiple trajectory points.

[0075] For example, in one instance, if the target dwell time corresponding to the intersection segment is greater than the preset dwell time, the effective dwell points and the target number of effective dwell points are determined from the multiple dwell points determined above.

[0076] Furthermore, in a specific example, the effective stopping points and the target number of effective stopping points can be obtained in the following manner; specifically, when the target stopping time corresponding to the intersection segment is greater than the preset stopping time, determining the effective stopping points and the target number of effective stopping points from multiple trajectory points (for example, step S205) as described above can specifically include: If the target dwell time corresponding to the intersection segment is greater than the preset dwell time, the dwell points within the preset time window corresponding to the intersection segment are determined as valid dwell points, so as to obtain the target number of valid dwell points.

[0077] For example, in one example, if the target dwell time for a certain intersection segment is 65 seconds, the total number of dwell points is 20, and the preset dwell time for the intersection segment is 60 seconds, then since the target dwell time is longer than the preset dwell time, all dwell points corresponding to the intersection segment can be considered as valid dwell points. In this case, the target number of valid dwell points is 20.

[0078] In other words, in this example, the target stopping time is calculated using trajectory points that exhibit stopping behavior, i.e., stopping points. At the same time, if the target stopping time is longer than the preset stopping time, the aforementioned stopping points are directly used as valid stopping points. This method is simple, efficient, and highly interpretable, thus providing data support for the subsequent rapid mining of traffic signal equipment.

[0079] It should be noted that, in practical applications, in order to further improve the information density and accuracy of "whether traffic signal equipment exists" carried by the target number of effective stopping points, and to avoid deviations caused by the sparse number, the solution of this application can count the effective stopping points within multiple preset time windows. For example, if the target stopping time corresponding to the intersection segment is greater than the preset stopping time, the stopping points within multiple preset time windows corresponding to the intersection segment are determined as effective stopping points. For example, all stopping points within multiple preset time windows that are used to calculate the target stopping time are considered as effective stopping points, and the total number of effective stopping points within these multiple preset time windows is taken as the target number. In this way, the prediction accuracy of the initial prediction information is further improved.

[0080] Thus, this disclosed solution provides a method for quickly obtaining effective stopping points. Specifically, when the target stopping time corresponding to an intersection segment exceeds a preset stopping time, the stopping point corresponding to that intersection segment is directly determined as a valid stopping point, and the number of valid stopping points is counted. This simplifies the process of determining valid stopping points, reduces the computational load of trajectory data processing, improves the calculation efficiency of valid stopping points, and ensures the accuracy and rationality of the statistics of valid stopping points. This provides a reliable data foundation for the initial prediction of traffic signal equipment at subsequent intersections, further improving the accuracy and robustness of traffic light recognition.

[0081] Furthermore, it should be noted that in practical applications, the preset stopping time can be determined based on the actual scenario of the target road. For example, the preset stopping time can be calculated using the following method: Identify designated intersections in the map data that already have traffic signal devices. Analyze the vehicle stopping characteristics at these intersections during each cycle (e.g., traffic light cycle), and use clustering to calculate the effective stopping duration. This effective stopping duration is then used as a reference value for a preset stopping duration. For example, the calculated effective stopping duration can be directly used as the preset stopping duration. Alternatively, the effective stopping durations of multiple designated intersections can be calculated, and the average of these effective stopping durations can be used as the preset stopping duration.

[0082] It should be noted that in practical applications, the distance between each vehicle and the traffic light varies, resulting in different stopping times. For example, the closer a vehicle is to the intersection, the shorter its stopping time; the farther away it is, the longer its stopping time. Therefore, when calculating the effective stopping time, the average stopping time can be considered. For instance, in one example, the stopping time of different vehicles at a designated intersection within a given timeframe can be statistically analyzed to obtain the average stopping time. This average stopping time can then be used as the effective stopping time.

[0083] Alternatively, we can further analyze the dwell time of different vehicles at the designated intersection within a certain period of time, and determine the actual distance of each vehicle from the designated intersection. We can then perform weighted processing based on the actual distance and dwell time, for example, by determining the weight of dwell time based on the actual distance, to obtain the weighted average dwell time at the designated intersection. In this case, the weighted average dwell time can be used as the effective dwell time.

[0084] It should be noted that the above is only an illustrative example. In actual applications, other methods can be used to obtain the preset dwell time, and this disclosure does not impose any specific restrictions on this.

[0085] Furthermore, it is understood that the preset stopping time corresponding to different target roads can be the same or different. For example, considering that the change cycle of traffic signal equipment in different areas may be different, adaptive adjustments can be made based on regional differences. This disclosed solution does not impose specific restrictions on this.

[0086] Step S206: Based on the target number of effective stopping points, obtain initial prediction information for the intersection segment. Here, the initial prediction information is used to indicate whether traffic signal equipment exists.

[0087] Step S207: Determine the target image data corresponding to the intersection segment, and verify the initial prediction information based on the target image data to obtain the target verification result.

[0088] Here, the target verification result is used to characterize the accuracy of the initial prediction information.

[0089] In this way, the proposed solution first determines whether there is a stopping behavior based on the time interval between trajectory points within the intersection and road segment, then uses the trajectory points with stopping behavior as stopping points, and finally determines the target stopping duration corresponding to the intersection and road segment based on the time interval between stopping points. Thus, by first identifying stopping behavior, then extracting stopping points, and finally determining the target stopping duration, the solution achieves standardization and accuracy in the calculation of target stopping duration. In the above process, trajectory points without stopping behavior are effectively eliminated, improving the reliability of stopping duration calculation and providing a stable and reliable data foundation for subsequent traffic signal equipment prediction.

[0090] Furthermore, in a specific example, the target dwell time corresponding to the intersection segment can be obtained in the following way; specifically, the above-mentioned determination of the target dwell time corresponding to the intersection segment based on the time interval between dwell points (for example, step S204) can specifically include: Step S204-1: Determine the total stopping time corresponding to the intersection segment based on the time interval between stopping points.

[0091] For example, in one instance, the time intervals between stopping points are summed to determine the total stopping time corresponding to the intersection segment.

[0092] It should be noted that the stopping points used in step S204-1 to calculate the total stopping time corresponding to the intersection segment can be stopping points within a single preset time window or stopping points within multiple preset time windows. For example, in one example, if the total stopping time determined based on stopping points within a single preset time window is greater than a specified threshold, then step S204-2 is executed; otherwise, if the total stopping time determined based on stopping points within a single preset time window is less than the specified threshold, then one or more stopping points within a single preset time window need to be added so that the total stopping time determined based on stopping points within multiple preset time windows is greater than the specified threshold. This provides favorable support for effectively improving the prediction accuracy of the initial prediction information. Alternatively, in another example, the number of stopping points within a single preset time window used to calculate the total stopping time can be counted, and if the counted number is less than another specified threshold, the number of stopping points within multiple preset time windows can be counted until the counted number meets the aforementioned threshold requirement.

[0093] Understandably, in another example, it is also possible to determine whether it is necessary to count the number of stops in multiple preset time windows based on two dimensions: total dwell time and number of stops. For example, if any one dimension does not meet the above requirements, it is necessary to count the number of stops in multiple preset time windows. This provides favorable support for improving the prediction accuracy of the initial prediction information.

[0094] Step S204-2: Based on the total dwell time corresponding to the intersection segment and the total number of dwell points corresponding to the intersection segment, determine the target dwell time corresponding to the intersection segment.

[0095] For example, in one example, we can first sum the time intervals between all adjacent stopping points of any single driving trajectory within the intersection segment to obtain the stopping time of the driving trajectory through the intersection segment; then, we can sum the stopping times of each driving trajectory within the intersection segment to obtain the total stopping time for the intersection segment.

[0096] For example, in one example, any intersection segment can contain multiple driving trajectories, such as p driving trajectories, denoted as driving trajectory T1, driving trajectory T2, ..., driving trajectory T... p Each driving trajectory may contain one or more time intervals between stopping points. For example, driving trajectory T1 contains N+1 stopping points and N time intervals, which can be denoted as time interval B1, time interval B2, ..., time interval B... N At this point, for the driving trajectory T1, the total dwell time corresponding to this intersection segment can be specifically recorded as: T 1B1 +T 1B2 +……+T 1BN Similarly, for a driving trajectory T2, if it contains M+1 stopping points and M time intervals, these M time intervals can be denoted as time interval B1, time interval B2, ..., time interval B... M For the driving trajectory T2, the total stopping time at this intersection segment can be specifically recorded as: T 2B1 +T 2B2 +……+T 2BM And so on, for the driving trajectory T p If there are q+1 stopping points and q time intervals, the q time intervals can be denoted as time interval B1, time interval B2, ..., time interval B. q For the driving trajectory T p Specifically, the total dwelling time corresponding to this intersection segment can be denoted as: T pB1 +T pB2 +……+T pBq Furthermore, the dwell times of each driving trajectory are summed to calculate the total dwell time corresponding to the intersection segment, denoted as T. The total dwell time T can then be expressed as: T=T 1B1 +T 1B2 +……+T 1BN +T 2B1 +T 2B2 +……+T 2BM +T pB1 +T pB2 +……+T pBq .

[0097] For example, in one scenario, there are five driving trajectories within a certain intersection segment: trajectory T1, trajectory T2, trajectory T3, trajectory T4, and trajectory T5. Trajectory T1 contains five time intervals between stopping points, which are 22s, 25s, 28s, 24s, and 26s, respectively, so the stopping duration of this trajectory is 125s. Trajectory T2 contains four time intervals between stopping points, which are 35s, 38s, 36s, and 41s, respectively, so the stopping duration of this trajectory is 150s. Trajectory T3 contains three time intervals between stopping points, which are 6s, 25s, 28s, 24s, and 26s, respectively, so the stopping duration of this trajectory is 150s. If the time intervals between the four stopping points in trajectory T4 are 0s, 65s, and 62s, then the total stopping time for this trajectory is 187s. Trajectory T4 contains four stopping points with time intervals of 78s, 82s, 75s, and 80s, so the total stopping time for this trajectory is 315s. Trajectory T5 contains four stopping points with time intervals of 105s, 98s, 102s, and 108s, so the total stopping time for this trajectory is 413s. Furthermore, by summing the stopping times of the five trajectories, the total stopping time for this intersection segment is 125 + 150 + 187 + 315 + 413 = 1300s.

[0098] Further, in one example, step S204-2 described above may specifically include: taking the ratio of the total dwell time corresponding to the intersection segment to the total number of dwell points corresponding to the intersection segment as the target dwell time corresponding to the intersection segment. In other words, the target dwell time at this time can be understood as the average dwell time.

[0099] For example, in one example, if the total dwell time corresponding to a certain intersection segment is T, and the total number of dwell points corresponding to the intersection segment is x, then the target dwell time can be specifically expressed as: T / x. For example, continuing with the above example, the total number of dwell points corresponding to the intersection segment can be expressed as x=[(N+1)+(M+1)+……+(q+1)], and the formula for calculating the target dwell time corresponding to the intersection segment can be expressed as T / [(N+1)+(M+1)+……+(q+1)].

[0100] Thus, this disclosed solution provides a specific method for obtaining the target dwell time, namely, first determining the total dwell time corresponding to the intersection segment based on the time interval between dwell points, and then obtaining the corresponding target dwell time based on the total dwell time corresponding to the intersection segment and the total number of dwell points. In this way, the information density of "whether traffic signal equipment exists" carried by the target dwell time is improved, providing reliable data support for the accurate prediction of traffic signal equipment at intersections in the future.

[0101] Figure 4This is an illustrative flowchart of a trajectory data processing method according to an embodiment of this application. Figure 3 This method can be optionally applied to electronic devices, such as personal computers, servers, and server clusters. It is understood that the above... Figures 1 to 3 The relevant content of the method shown in any of the attached figures can also be applied to this example, and the relevant content will not be described again in this example.

[0102] Furthermore, the method includes at least a portion of the following: For example... Figure 4 As shown, it includes: Step S401: Based on multiple driving trajectories of the vehicle, determine the intersections and road segments in the target road.

[0103] Here, each driving trajectory contains multiple trajectory points.

[0104] Step S402: Determine the target dwell time corresponding to the intersection segment based on the time interval between the trajectory points corresponding to the intersection segment.

[0105] Step S403: If the target dwell time corresponding to the intersection segment is greater than the preset dwell time, determine the effective dwell points and the number of effective dwell points from multiple trajectory points.

[0106] For details regarding time intervals, stopping points, target stopping durations, and preset stopping durations, please refer to the examples above; they will not be repeated here.

[0107] Step S404: When the number of target valid stopping points is greater than the preset number of stopping points, obtain the initial prediction information indicating the possible presence of traffic signal equipment at the intersection.

[0108] In another example, if the target dwell time corresponding to the intersection segment is less than or equal to the preset dwell time, it is determined that there is no traffic signal equipment in the intersection segment. At this time, the initial prediction information indicating that there is no traffic signal equipment in the intersection segment can be obtained.

[0109] Additionally, it should be noted that the preset number of stopping points can be set according to the actual application scenario. For example, in one example, the preset number of stopping points can be a fixed threshold, such as 20, 30, or 50; or, in another example, the preset number of stopping points can also be adaptively determined based on historical statistical data, for example, by analyzing the distribution of the effective number of stopping points at intersections with known traffic signal equipment, and selecting its mean or median as the preset number of stopping points, etc. This disclosure does not impose specific restrictions on this.

[0110] Furthermore, to improve the accuracy of the judgment results, in one example, the preset number of stopping points can be dynamically adjusted according to time period, road grade, or traffic flow. For example, in a specific example, during peak traffic hours, the preset number of stopping points threshold can be appropriately increased to reduce the occurrence of false judgments; during off-peak traffic hours, the preset number of stopping points threshold can be decreased to improve detection sensitivity. This effectively enhances the practicality, applicability, and robustness of the disclosed solution.

[0111] Step S405: If the initial prediction information indicates that there is a traffic signal device at the intersection, determine whether there is an environmental image with a time later than the stop point time. For example, determine whether there is a latest environmental image with a time later than the stop point time. If there is, proceed to step S406; otherwise, that is, if there is no environmental image with a time later than the stop point time, proceed to step S407.

[0112] It should be noted that the dwell time can be the acquisition timestamp corresponding to a valid dwell point. Furthermore, to obtain the latest environmental image, this dwell time can be specifically the acquisition timestamp of the latest dwell point among all valid dwell points.

[0113] Furthermore, the environmental images mentioned above can specifically refer to image data containing visual information of intersections and road segments collected by the acquisition device during vehicle operation; or, image data containing visual information of intersections and road segments collected by other road testing equipment. It should be noted that this disclosure does not impose specific restrictions on the acquisition method of the environmental images.

[0114] Step S406: Use the environmental image as the target image data. Proceed to step S408.

[0115] It should be noted that the environmental image can specifically refer to image data reflecting the traffic environment around the intersection / road segment. For example, in one instance, when multiple environmental images meet the conditions, the target image data can be determined using at least one of the following methods: selecting the environmental image whose time is closest to the stopping point time; selecting the environmental image with the highest clarity; or selecting multiple environmental images with different perspectives and stitching them together as the target image data.

[0116] In other words, in this example, when the initial prediction information indicates that there is a traffic signal device at the intersection, historical images of that intersection are acquired. For example, the latest environmental image of that intersection with a time later than the stopping point time is acquired. In this way, the accuracy of the initial prediction information is verified by combining the environmental image and using visual recognition. This effectively avoids the time and computing resource consumption caused by repeatedly collecting on-site images. At the same time, it realizes the fusion and verification of trajectory data and image data. While ensuring the processing effect, it improves the processing efficiency, thereby laying the foundation for improving the accuracy and reliability of traffic signal device identification.

[0117] Step S407: Trigger the generation of acquisition command to acquire target image data. Proceed to step S408.

[0118] Here, the acquisition command is used to instruct the acquisition of target image data corresponding to the intersection and road segment. For example, in one example, when there is no environmental image that meets the time conditions in the existing image data, or when there is no environmental image for the intersection and road segment, the acquisition command is triggered. In this way, the image acquisition process can be started in a timely manner, avoiding the interruption of the subsequent initial prediction information verification process due to missing images, thereby ensuring the continuity and integrity of traffic signal equipment verification.

[0119] Furthermore, the acquisition command can be a control signal used to control the image acquisition device to acquire target image data corresponding to the intersection or road segment. For example, in one example, the acquisition command can be sent to at least one image acquisition device, including but not limited to: vehicle-mounted camera equipment, road monitoring camera equipment, inspection equipment, or drone acquisition equipment.

[0120] In other words, in this example, when the initial prediction information indicates that there is a traffic signal device at the intersection, and historical images of the intersection cannot be obtained, an image acquisition command is generated to acquire images. After the acquisition is completed, the images are combined with environmental images, and the accuracy of the initial prediction information is verified through visual recognition. In other words, this disclosed solution provides an efficient resource processing mode of on-demand triggering and layered verification. This mode precisely controls the timing of the high-cost data acquisition and processing process, avoiding the blockage of the verification process due to missing image data. At the same time, it avoids the time and computing resource consumption caused by repeated on-site image acquisition, saving computing and storage resources, and ensuring the continuity and feasibility of subsequent initial prediction information verification. Thus, it achieves refined and economical use of system resources while ensuring that key information can be verified.

[0121] Step S408: Based on the target image data, verify the initial prediction information to obtain a target verification result. Here, the target verification result is used to characterize the accuracy of the initial prediction information.

[0122] For example, in one instance, image recognition processing can be performed on the target image data to verify the initial prediction information. For instance, a target detection algorithm can be used to identify whether traffic signal equipment exists in the target image data to verify the initial prediction information.

[0123] Specifically, image recognition processing is performed on the target image data. If the recognition result indicates the presence of traffic signal equipment in the target image data, the recognition result is consistent with the initial prediction information verification, and the target verification result is "verification passed." If the recognition result indicates the absence of traffic signal equipment in the target image data, the recognition result is inconsistent with the initial prediction information verification, and the target verification result is "verification failed." Alternatively, other target image data corresponding to the intersection segment can be acquired to re-identify whether traffic signal equipment exists.

[0124] For example, in one instance, for a certain intersection segment, the initial prediction information is "suspected presence of traffic signal equipment". Further, based on the stop time (e.g., 10:05:20), two environmental images with capture times of 10:05:25 and 10:05:40 are selected from the vehicle's image data, and the frame with higher clarity is chosen as the target image data. Further, image recognition processing is performed on the target image data. If the recognition result indicates the presence of traffic signal equipment in the target image data, the recognition result is consistent with the initial prediction information, and the target verification result is output as "verification passed".

[0125] Alternatively, in another example, for a certain intersection segment, the initial prediction information is "suspected presence of traffic signal equipment," but no image data is matched at the corresponding stop time. In this case, a collection command for the intersection segment is triggered, and the roadside camera corresponding to the intersection segment is controlled to perform a real-time image collection step to collect the target image data corresponding to the intersection segment. Further, the target image data is acquired and subjected to image recognition processing. If the recognition result indicates that traffic signal equipment exists in the target image data, the recognition result is consistent with the initial prediction information, and the target verification result is output as "verification passed."

[0126] In this way, the proposed solution compares the target number of effective stopping points with the preset number of stopping points, and when the target number is greater than the preset number of stopping points, obtains the initial prediction information of the intersection section where traffic signal equipment is suspected. Thus, it can make a preliminary judgment on whether there is traffic signal equipment at the intersection section based on the number of effective stopping points, providing reliable prior information for subsequent verification steps based on image data, thereby improving the overall accuracy and robustness of the identification.

[0127] Furthermore, in a specific example, after obtaining the target verification result, the current map data can be optimized based on the target verification result. Specifically, in one example, the presence of traffic signal equipment in the intersection / road segment can be determined based on the number of effective stopping points corresponding to the intersection / road segment and the target image data corresponding to the intersection / road segment, and then the current map data can be updated based on the determination result.

[0128] For example, when the number of valid stopping points within one or more preset time windows exceeds the preset number of stopping points, the stopping characteristics of the intersection segment can be considered valid. Based on this, the following judgment can be performed: (1) When the number of effective stopping points is greater than the number of preset stopping points, and there is no corresponding traffic signal equipment in the target image data, it is determined that the intersection section needs to add traffic signal equipment. Furthermore, the addition result is output for the planning department, traffic management department or road maintenance unit to make decisions and construction arrangements, so as to achieve accurate, dynamic supplementation and optimization of traffic signal equipment.

[0129] (2) When the number of effective stopping points is less than the preset number of stopping points, and there is a corresponding traffic signal device in the target image data, the modification result is output and the map data is updated based on the modification result.

[0130] Specifically, the modification result can represent at least one of the following: The traffic signal equipment is redundant. The traffic signal equipment is in an inactive state, for example, in a blackout state; The traffic signal equipment is in a warning state, such as a flashing yellow light.

[0131] It should be noted that traffic signal devices in a warning state are usually used to prompt vehicles to slow down, but do not require them to stop; traffic signal devices in an inactive state usually do not require vehicles to stop, and therefore will not cause vehicles to stop noticeably.

[0132] By using the above methods, traffic signal equipment can be dynamically updated based on the joint analysis of vehicle trajectory data and target image data, thereby improving the accuracy and timeliness of map information.

[0133] Figure 5 This is a schematic diagram illustrating the result of trajectory data processing according to an embodiment of this application. For example... Figure 5 As shown in the figure, this diagram illustrates the spatial and temporal distribution of trajectory points as a vehicle approaches an intersection. For example, Figure 5 As shown, the horizontal axis represents the distance of the vehicle from the intersection or traffic signal equipment (unit: meters (m)), and the vertical axis represents the time corresponding to the trajectory point (unit: seconds (s)). Furthermore, the multiple green dots in the figure represent trajectory points collected during the vehicle's journey. These trajectory points generally show a distribution trend from far to near, reflecting the vehicle's gradual approach to the intersection. Further, in areas near the intersection, some trajectory points exhibit a clear clustering phenomenon, forming local stacking or curved distributions in the time dimension, for example... Figure 5 As shown by the red curved line segment, this type of trajectory point reflects the behavior of the corresponding vehicle when it is stationary or moving at low speed at that location.

[0134] Furthermore, such as Figure 5 As shown in the blue box area of ​​the figure, multiple stopping behaviors can be observed repeatedly occurring at similar spatial locations, exhibiting a certain regularity in terms of time intervals. This regularity corresponds to the periodic control of the traffic signal equipment. For example, in a specific example, the period of the traffic signal equipment is approximately 120 seconds, and the time intervals between different stopping behaviors roughly conform to this periodic characteristic. Furthermore, within the red box area of ​​the figure, statistical analysis of stopping behaviors can be performed, such as analyzing characteristic parameters like the number of stops per unit time and the average stopping duration. By analyzing these characteristic parameters, the stopping characteristics of this intersection segment can be obtained.

[0135] Furthermore, based on the spatial clustering and temporal periodicity of the aforementioned stopping behavior, it can be determined whether traffic signal equipment exists at the intersection. Specifically, when the cumulative number of effective stopping points exceeds a preset threshold within one or more time windows, the stopping characteristics at that location can be considered significant, thereby generating corresponding initial prediction information.

[0136] Furthermore, the status of the intersection / road segment can be determined by combining existing traffic signal equipment information on the map. For example, if the cumulative number of valid stopping points exceeds a threshold and no traffic signal equipment exists on the map, it can be determined that a new traffic signal equipment needs to be added at that location. If the cumulative number of valid stopping points is less than a threshold and traffic signal equipment already exists on the map, it can be determined that the traffic signal equipment may be in an abnormal state, such as redundant, not activated (black light), or flashing yellow. In this way, stopping behavior characteristics in vehicle trajectory data can be used to identify and update the status of traffic signal equipment at intersections / road segments, thereby improving the accuracy and real-time performance of map data.

[0137] Specifically, Figure 6 This is a schematic diagram illustrating the implementation flow of a trajectory data processing method according to an embodiment of this application in a specific example. The trajectory data processing method in this example can be implemented through the following steps, specifically including: Step S601: Filter driving trajectories.

[0138] For example, in one instance, based on the vehicle type labels carried in the trajectory data, non-target types of driving trajectories such as walking and cycling can be removed, while the driving trajectories of specified vehicles (such as small passenger cars) can be retained, to ensure that the trajectory analysis focuses on the main vehicle types controlled by the traffic signal equipment at the intersection.

[0139] Step S602: Match the multiple driving trajectories of the vehicle with the preset road network data to determine the intersection segment.

[0140] Furthermore, the intersection segment can be determined in the following manner; specifically, step S602 can include: Step S602-1: Match the multiple driving trajectories of the vehicle with the preset road network data to determine the actual road to which each driving trajectory data belongs.

[0141] Here, the preset road network data refers to road vector data that has been pre-collected and standardized, such as map road network data for a specified area.

[0142] Furthermore, after determining the preset road network data, matching can be performed in the following ways, such as, but not limited to, one or more of spatial projection matching, topological path matching, and spatial nearest neighbor matching. In this way, the positioning information of the trajectory points is mapped to the preset road network data, thereby accurately locating the actual road corresponding to each driving trajectory, which provides favorable support for the subsequent accurate discovery of traffic signal equipment at intersections and road sections.

[0143] Step S602-2: Based on the number of trajectory points in the actual road, filter out the actual roads with a trajectory coverage rate higher than the preset coverage rate threshold as target roads.

[0144] It should be noted that the preset coverage threshold is a critical value pre-set based on data calculation requirements and analysis accuracy requirements, and can be set according to actual application conditions. This disclosure does not impose specific restrictions on it.

[0145] It should also be noted that the trajectory coverage rate is an indicator that quantitatively represents the degree to which vehicle trajectories cover the actual road. For example, it can be specifically represented by the total number of trajectory points contained within a unit length of the actual road. In this case, the higher the trajectory coverage rate, the more trajectory points there are on the actual road, which makes it easier to accurately mine the data. This can effectively avoid the problem of inaccurate feature judgment caused by a small number of trajectories, and provide data support for the subsequent accurate mining of traffic signal equipment at intersections and road sections and the improvement of the accuracy of prediction results.

[0146] For example, in one scenario, differentiated trajectory statistics strategies can be employed for road segments with different trajectory access volumes (i.e., traffic) to avoid biases caused by insufficient trajectory point data. Specifically, for example, for trajectory segments with high trajectory access volumes, single-day trajectory data can be used directly to meet the coverage calculation requirements; for trajectory segments with low trajectory access volumes, accumulated trajectory data from multiple days is used for statistical analysis to compensate for the data volume bias and prevent distorted coverage judgments and improper target road selection due to insufficient trajectory samples.

[0147] Step S602-3: Determine the intersection segment based on the intersection location in the target road.

[0148] Here, the length of the intersection segment is shorter than the length of the target road. For example, a segment 100 meters away from the intersection on the target road can be considered the intersection segment.

[0149] Step S603: Determine the stopping point.

[0150] It should be noted that the stopping point can be a trajectory point that exhibits stopping behavior within a preset time window. For example, in one example, the presence of stopping behavior at a trajectory point can be determined based on the time interval between trajectory points, a preset time threshold, and the coordinate information corresponding to the trajectory point. Further, within the preset time window, if the coordinate information corresponding to two adjacent trajectory points is greater than a preset distance threshold (e.g., A meters), and the time interval between the trajectory points is greater than a preset time threshold (e.g., B seconds), then it is determined that the two adjacent trajectory points exhibit stopping behavior within the preset time window; otherwise, it is determined that the two adjacent trajectory points do not exhibit stopping behavior within the preset time window.

[0151] Furthermore, for two adjacent trajectory points that exhibit stopping behavior within a preset time window, such as the i-th trajectory point and the (i+1)-th trajectory point, both the i-th trajectory point and the (i+1)-th trajectory point can be considered as stopping points. Further, all driving trajectories within any intersection segment are traversed to obtain multiple stopping points corresponding to that intersection segment.

[0152] Step S604: Set the effective dwell time (i.e. the aforementioned preset dwell time).

[0153] Here, the effective stopping time is a pre-set time threshold. For example, it can be a reference true value obtained by statistically analyzing and clustering the actual stopping behavior of most vehicles in each traffic light cycle by filtering intersections and road segments with marked traffic lights in the map data. It should be noted that the effective stopping time is an empirical value and can be set according to the actual scenario. This public solution does not impose specific restrictions on it.

[0154] It should be noted that the dwell time varies depending on the distance between different vehicles and the intersection. For example, vehicles closer to the intersection can pass through within the current signal cycle, resulting in a relatively short dwell time; while vehicles farther from the intersection need to wait for the next signal cycle, resulting in a relatively long dwell time. Therefore, different effective dwell times can be used as statistical thresholds to improve the accuracy and reliability of the final effective dwell time.

[0155] Step S605: Calculate the average dwell time (i.e., the aforementioned target dwell time).

[0156] Furthermore, the average dwell time can be calculated using the following steps, specifically, step S605 may include: Step S605-1: Determine the total stopping time corresponding to the intersection segment based on the time interval between stopping points.

[0157] Furthermore, in one example, based on a single driving trajectory of an intersection segment, the time intervals between all adjacent stopping points of the driving trajectory within the intersection segment can be summed to obtain the stopping time of the driving trajectory through the intersection segment; then, the stopping times of each driving trajectory within the intersection segment can be summed to obtain the total stopping time for the intersection segment.

[0158] Step S605-2: Based on the total dwell time corresponding to the intersection segment and the total number of dwell points corresponding to the intersection segment, determine the average dwell time corresponding to the intersection segment.

[0159] For example, in one example, the ratio of the total dwell time corresponding to the intersection segment to the total number of dwell points corresponding to the intersection segment (that is, the total number of dwell points used to calculate the total dwell time) is used as the average dwell time corresponding to the intersection segment.

[0160] Step S606: Determine whether the average dwell time is greater than the effective dwell time. If yes, proceed to step S607; otherwise, if the average dwell time is less than or equal to the effective dwell time, proceed to step S608.

[0161] Step S607: Determine the stopping points corresponding to the intersection and road segment as valid stopping points to obtain the target number of valid stopping points within the preset time window. Proceed to step S609.

[0162] For example, in one instance, if the average dwell time corresponding to the intersection segment is greater than the preset dwell time, the dwell point corresponding to the intersection segment is determined as a valid dwell point, thereby obtaining the target number of valid dwell points for subsequent existence analysis of traffic signal equipment.

[0163] Step S608: Determine the stopping points corresponding to the intersection and road segment as invalid stopping points.

[0164] For example, in one instance, if the average dwell time corresponding to the intersection segment is less than or equal to the preset dwell time, then the dwelling behavior at the intersection segment is determined to be normal slow traffic or temporary stopping, and the dwelling point corresponding to the intersection segment is determined to be an invalid dwelling point, so as to avoid interfering with the existence analysis of subsequent traffic signal equipment.

[0165] Step S609: Based on the target number of effective stopping points, obtain initial prediction information for the intersection segment.

[0166] For example, in one example, when the target number of valid stopping points is greater than the preset number of stopping points, initial prediction information indicating the possible presence of traffic signal equipment at the intersection is obtained. Alternatively, in another example, when the target number of valid stopping points is less than or equal to the preset number of stopping points, initial prediction information indicating the absence of traffic signal equipment at the intersection is obtained.

[0167] Furthermore, the status of traffic signal equipment at the intersection can be updated by combining the existing status information of the traffic signal equipment in the map data.

[0168] For example, in one specific example, if the target number of valid stopping points is greater than the preset number of stopping points, and there are no traffic signal devices in the map data, then it is determined that traffic signal devices need to be added to the intersection segment, and the update result is "Need to add". Alternatively, in another specific example, if the target number of valid stopping points is less than or equal to the preset number of stopping points, and there are traffic signal devices in the map data, then it is determined that traffic signal devices need to be modified for the intersection segment, and the update result is "Need to modify".

[0169] Step S610: If the initial prediction information indicates that there is a traffic signal device at the intersection, determine whether there is an environmental image with a time later than the stop point time. If yes, proceed to step S611; otherwise, if there is no environmental image with a time later than the stop point time, proceed to step S612.

[0170] Step S611: Use the environmental image as the target image data. Proceed to step S613.

[0171] For example, in one instance, when there are multiple environmental images that meet the conditions, the target image data can be determined by at least one of the following methods: selecting the environmental image whose time is closest to the dwell time; selecting the environmental image with the highest clarity; or selecting multiple environmental images with different perspectives and stitching them together as the target image data.

[0172] Step S612: Issue the data acquisition command. Proceed to step S611.

[0173] Here, the acquisition command is used to instruct the acquisition of target image data corresponding to the intersection and road segment. For example, in one example, when there is no environmental image that meets the time conditions in the existing image data, or when there is no environmental image for the intersection and road segment, the acquisition command is triggered. In this way, the image acquisition process can be started in a timely manner, avoiding the interruption of the subsequent initial prediction information verification process due to missing images, thereby ensuring the continuity and integrity of traffic signal equipment verification.

[0174] Furthermore, the acquisition command can be a control signal used to control the image acquisition device to acquire target image data corresponding to the intersection or road segment. For example, in one example, the acquisition command can be sent to at least one image acquisition device, including but not limited to: vehicle-mounted camera equipment, road monitoring camera equipment, inspection equipment, or drone acquisition equipment.

[0175] Step S613: Based on the target image data, perform a second verification on the initial prediction information to obtain the target verification result.

[0176] For example, in a specific example, for a certain intersection segment, the initial prediction information is "suspected traffic signal equipment exists", but no image data is matched at the corresponding stop time. At this time, a collection instruction for the intersection segment is triggered, and the roadside camera equipment corresponding to the intersection segment is controlled to perform a real-time image collection step to collect the target image data corresponding to the intersection segment. The target image data is then processed by image recognition. If the recognition result is that a traffic signal equipment exists in the target image data, the recognition result is consistent with the initial prediction information, and the target verification result is output as "verification passed".

[0177] Step S614: Manual verification and map data update.

[0178] For example, in one instance, the target verification result obtained in step S613 is submitted for manual review. This review is conducted by professionals who combine the actual road conditions, on-site photos, and surveillance videos from multiple dimensions to verify and confirm the information. Specifically, after manual verification, the status of traffic signal equipment in the map database is updated based on the judgment result. For instance, for intersections where the judgment result is "needs to add" traffic signal equipment, the traffic light label information for that intersection is added to the map database; for intersections where the judgment result is "needs to modify" traffic signal equipment, the status of the traffic lights (e.g., normal, faulty, removed) is updated in the map database, or expired label information is deleted. This ensures the accuracy and timeliness of traffic signal equipment information for intersections in the map database, providing reliable data support for scenarios such as map navigation and traffic condition warnings.

[0179] This disclosure also provides a trajectory data processing device, such as... Figure 7 As shown, it includes: The trajectory data processing unit 701 is used to determine the intersection segment in the target road based on multiple driving trajectories of the vehicle; wherein each driving trajectory contains multiple trajectory points; based on the time interval between the trajectory points corresponding to the intersection segment, determine the target stopping time corresponding to the intersection segment; if the target stopping time corresponding to the intersection segment is greater than a preset stopping time, determine the effective stopping points and the target number of effective stopping points from the multiple trajectory points; based on the target number of effective stopping points, obtain initial prediction information for the intersection segment, wherein the initial prediction information is used to indicate whether traffic signal equipment exists; The prediction result verification unit 702 is used to determine the target image data corresponding to the intersection segment, and to verify the initial prediction information based on the target image data to obtain the target verification result; wherein, the target verification result is used to characterize the accuracy of the initial prediction information.

[0180] In a specific example of the scheme disclosed herein, the trajectory data processing unit is specifically used for: Based on the time interval between the trajectory points corresponding to the intersection and road segment, determine whether the trajectory points have stopped behavior; Trajectory points that exhibit stopping behavior within a preset time window will be used as stopping points to obtain multiple stopping points corresponding to the intersection and road segment; Based on the time interval between stopping points, the target stopping duration corresponding to the intersection and road segment is determined.

[0181] In a specific example of the scheme disclosed herein, the trajectory data processing unit is specifically used for: Based on the time interval between stopping points, the total stopping time corresponding to the intersection and road segment is determined; Based on the total dwell time corresponding to the intersection and road segment, and the total number of dwell points corresponding to the intersection and road segment, the target dwell time corresponding to the intersection and road segment is determined.

[0182] In a specific example of the scheme disclosed herein, the trajectory data processing unit is specifically used for: If the target dwell time corresponding to the intersection segment is greater than the preset dwell time, the dwell points within the preset time window corresponding to the intersection segment are determined as valid dwell points, so as to obtain the target number of valid dwell points.

[0183] In a specific example of the scheme disclosed herein, the trajectory data processing unit is specifically used for: When the number of target effective stopping points is greater than the preset number of stopping points, initial prediction information is obtained to indicate the presence of traffic signal equipment at the intersection.

[0184] The prediction result verification unit is specifically used for: Given that the initial prediction information indicates that there may be traffic signal equipment at the intersection, determine whether there are environmental images with a time later than the stopping point time. If an environmental image of the intersection segment is found to exist at a time later than the stopping point time, the environmental image is used as the target image data.

[0185] In a specific example of the scheme disclosed herein, the prediction result verification unit is further configured to: If it is determined that there is no environmental image of the intersection segment with a time greater than or equal to the dwell time, a data acquisition command is triggered, wherein the data acquisition command is used to instruct the acquisition of target image data corresponding to the intersection segment.

[0186] In a specific example of the scheme disclosed herein, the trajectory data processing unit is specifically used for: The vehicle's multiple driving trajectories are matched with preset road network data to determine the actual road to which each driving trajectory data belongs; Based on the number of trajectory points in the actual road, actual roads with a trajectory coverage rate higher than a preset coverage threshold are selected as target roads; Based on the location of intersections in the target road, the intersection segments are determined, wherein the length of the intersection segments is shorter than the length of the target road.

[0187] For a description of the specific functions and examples of each unit of the apparatus in this disclosure embodiment, please refer to the relevant descriptions of the corresponding steps in the above method embodiments, which will not be repeated here.

[0188] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0189] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0190] Figure 8 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0191] like Figure 8 As shown, device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 802 or a computer program loaded from storage unit 808 into random access memory (RAM) 803. RAM 803 may also store various programs and data required for the operation of device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via bus 804. Input / output (I / O) interface 805 is also connected to bus 804.

[0192] Multiple components in device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of monitors, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0193] The computing unit 801 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as trajectory data processing methods. For example, in some embodiments, the trajectory data processing method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the trajectory data processing method described above may be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to perform trajectory data processing methods by any other suitable means (e.g., by means of firmware).

[0194] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0195] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0196] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0197] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0198] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0199] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0200] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0201] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A trajectory data processing method, comprising: Based on multiple driving trajectories of the vehicle, the intersections and road segments in the target road are identified; each driving trajectory contains multiple trajectory points. Based on the time interval between the trajectory points corresponding to the intersection and road segment, the target dwell time corresponding to the intersection and road segment is determined; If the target dwell time corresponding to the intersection segment is greater than the preset dwell time, determine the effective dwell points and the target number of effective dwell points from multiple trajectory points; Based on the target number of effective stopping points, initial prediction information is obtained for the intersection segment, wherein the initial prediction information is used to indicate whether traffic signal equipment exists; The target image data corresponding to the intersection segment is determined, and the initial prediction information is verified based on the target image data to obtain the target verification result; wherein, the target verification result is used to characterize the accuracy of the initial prediction information.

2. The method according to claim 1, wherein, The determination of the target dwell time corresponding to the intersection segment based on the time interval between the trajectory points corresponding to the intersection segment includes: Based on the time interval between the trajectory points corresponding to the intersection and road segment, determine whether the trajectory points have stopped behavior; Trajectory points that exhibit stopping behavior within a preset time window will be used as stopping points to obtain multiple stopping points corresponding to the intersection and road segment; Based on the time interval between stopping points, the target stopping duration corresponding to the intersection and road segment is determined.

3. The method according to claim 2, wherein, The determination of the target stopping duration for the intersection segment based on the time interval between stopping points includes: Based on the time interval between stopping points, the total stopping time corresponding to the intersection and road segment is determined; Based on the total dwell time corresponding to the intersection and road segment, and the total number of dwell points corresponding to the intersection and road segment, the target dwell time corresponding to the intersection and road segment is determined.

4. The method according to claim 2 or 3, wherein, When the target dwell time corresponding to the intersection segment is greater than the preset dwell time, the effective dwell points and the number of effective dwell points are determined from multiple trajectory points, including: If the target dwell time corresponding to the intersection segment is greater than the preset dwell time, the dwell points within the preset time window corresponding to the intersection segment are determined as valid dwell points, so as to obtain the target number of valid dwell points.

5. The method according to claim 2 or 3, wherein, The initial prediction information for the intersection segment is obtained based on the target number of effective stopping points, including: When the number of target effective stopping points is greater than the preset number of stopping points, initial prediction information is obtained to indicate the presence of traffic signal equipment at the intersection.

6. The method according to claim 5, wherein, The determination of the target image data corresponding to the intersection segment includes: Given that the initial prediction information indicates that there may be traffic signal equipment at the intersection, determine whether there are environmental images with a time later than the stopping point time. If an environmental image of the intersection segment is found to exist at a time later than the stopping point time, the environmental image is used as the target image data.

7. The method according to claim 6, further comprising: If it is determined that there is no environmental image of the intersection segment with a time greater than or equal to the dwell time, a data acquisition command is triggered, wherein the data acquisition command is used to instruct the acquisition of target image data corresponding to the intersection segment.

8. The method according to claim 1, wherein, The method of determining intersections and road segments in the target road based on multiple driving trajectories of the vehicle includes: The vehicle's multiple driving trajectories are matched with preset road network data to determine the actual road to which each driving trajectory data belongs; Based on the number of trajectory points in the actual road, actual roads with a trajectory coverage rate higher than a preset coverage threshold are selected as target roads; Based on the location of intersections in the target road, the intersection segments are determined, wherein the length of the intersection segments is shorter than the length of the target road.

9. A trajectory data processing device, comprising: The trajectory data processing unit is used to determine the intersection segment in the target road based on multiple driving trajectories of the vehicle; wherein each driving trajectory contains multiple trajectory points; based on the time interval between the trajectory points corresponding to the intersection segment, determine the target stopping duration corresponding to the intersection segment; if the target stopping duration corresponding to the intersection segment is greater than a preset stopping duration, determine the effective stopping points and the target number of effective stopping points from the multiple trajectory points; based on the target number of effective stopping points, obtain initial prediction information for the intersection segment, wherein the initial prediction information is used to indicate whether traffic signal equipment exists; The prediction result verification unit is used to determine the target image data corresponding to the intersection segment, and to verify the initial prediction information based on the target image data to obtain the target verification result; wherein, the target verification result is used to characterize the accuracy of the initial prediction information.

10. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-8.

11. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-8.

12. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-8.

Images