Map data acquisition system and method

Abstract

The invention relates to the technical field of electronic maps, and discloses a map data acquisition system and method. The system comprises an acquisition vehicle generation module, an acquisition task issuing module, a vehicle end acquisition module and a task tracking module. The collection vehicle generation module screens high-frequency passing vehicles and tracks through track clustering; the acquisition task issuing module segments the track to generate a strategy containing a vehicle code, a period and a plurality of segments of acquisition areas; the vehicle end acquisition module judges an acquisition area according to the longitude and latitude and the distance between the starting point and the ending point in a period, stores point cloud and image data, and uploads the data in a slicing manner; and the task tracking module monitors data uploading. According to the method, logistics vehicles and accurate positioning areas are reused, real-time slice uploading is realized, the cost is reduced, the period is shortened, the data freshness and accuracy are improved, and the method is suitable for high-precision map acquisition and updating of logistics scenes.

Description

Technical Field

[0001] This invention relates to the field of electronic map technology, specifically to a map data acquisition system and method, which is based on logistics operation vehicles and vehicle-cloud collaboration, and is particularly suitable for scenarios where high-precision map data can be efficiently acquired, uploaded in real time, and dynamically updated by reusing logistics operation vehicles. Background Technology

[0002] With the rapid development of fields such as intelligent driving and smart transportation, high-definition (HD) maps, as core foundational data, are crucial to system performance due to their freshness and accuracy. Currently, the collection of HD map data mainly relies on dedicated collection vehicles. These vehicles need to be equipped with expensive hardware such as LiDAR and high-precision positioning equipment, and require dedicated scheduling and operation, resulting in high data collection costs and significant pressure on vehicle resource reserves.

[0003] Meanwhile, traditional data collection methods suffer from long data collection cycles: dedicated vehicles need to traverse the collection area along fixed routes, and data is often uploaded in batches after collection, which can easily affect map update efficiency due to transmission delays; moreover, the coordination between the distribution of collection tasks and the execution on the vehicle is insufficient, making it difficult to quickly respond to dynamic road condition changes.

[0004] Furthermore, accurate positioning of the data collection area remains a pain point in the industry. Existing technologies sometimes identify the start and end points of the pre-collected roads, triggering a start marker when the vehicle enters the start point and an end marker when it reaches the end point. This allows for the extraction of target road segment data from the collected data to define the collection area. However, this approach still has room for improvement in terms of dynamic adaptability to the collection area and the collaborative mechanism with the data collection platform. Other related technologies either focus on extracting static interest surface coordinates or emphasize dynamic map updates for autonomous vehicles. Neither has developed an effective regional positioning solution for HD map data collection in logistics scenarios. This often leads to data redundancy or omissions in practical applications due to vehicle trajectory deviations and blurred area boundaries, increasing the complexity of subsequent data processing.

[0005] Therefore, how to reduce data collection costs, shorten the data collection cycle, and achieve more accurate and efficient data collection area positioning in logistics scenarios has become a key issue that urgently needs to be addressed in the field of HD map data collection. Summary of the Invention

[0006] The purpose of this invention is to provide a map data acquisition system, comprising: a vehicle generation module for acquiring vehicle and trajectory data within an electronic fence of a historical acquisition area, performing trajectory clustering based on a threshold, and determining the input data for the acquisition task; an acquisition task distribution module for generating an acquisition task strategy based on the acquisition task input data and distributing it to the vehicle terminal; a vehicle terminal acquisition module for acquiring data based on the acquisition task strategy and the relationship between the current vehicle terminal's latitude and longitude, start point, and end point within the acquisition task period; and a task tracking module for querying the data uploaded by the vehicle terminal based on the vehicle code and the acquisition task period.

[0007] According to one embodiment of the present invention, the vehicle generation module is used to determine that multiple trajectories of the vehicle are similar when the heading angle of the vehicle is less than or equal to a first angle threshold and the spatial distance of the vehicle is less than or equal to a first length threshold, and to group the similar trajectories into the same cluster.

[0008] According to one embodiment of the present invention, the vehicle generation module is used to acquire the vehicle and trajectory data that appear most frequently during the historical period as input data for the data acquisition task.

[0009] According to one embodiment of the present invention, the data acquisition task distribution module is used for: Based on the vehicle code and trajectory data generated by the vehicle generation module, the trajectory is interrupted at a first predetermined distance to generate multiple segment trajectories. A start point and an end point are taken in each segment trajectory, wherein the start point is the first trajectory point of the segment trajectory and the end point is the last trajectory point of the segment trajectory. The task distribution module is used to generate a collection task strategy based on the vehicle code, the collection task cycle, and multiple collection areas consisting of the start point and the end point.

[0010] According to one embodiment of the present invention, the vehicle-mounted data acquisition module is used for: Confirm that the current time is within the data collection task cycle, and confirm that the relationship between the current vehicle latitude and longitude and the starting point and the ending point meets preset conditions. Store point cloud data and image data for data acquisition. Upload the collected point cloud data and image data.

[0011] According to one embodiment of the present invention, the vehicle-mounted data acquisition module is used to confirm that the straight-line distance between the current vehicle-mounted latitude and longitude and the starting point and the straight-line distance between the current vehicle-mounted latitude and longitude and the ending point are both less than the straight-line distance between the starting point and the ending point, thus confirming that a preset condition is met.

[0012] According to one embodiment of the present invention, the vehicle-mounted acquisition module is used to slice the acquired point cloud data and image data at predetermined unit time intervals and upload the acquired point cloud data and image data in a slice mode.

[0013] According to one embodiment of the present invention, the map data acquisition system further includes a data storage module for storing the vehicle code, driving date and trajectory data of each vehicle; It is also used to store the electronic fences of all areas that need to be collected, their corresponding unique IDs, and to receive and store the data collected by the vehicle.

[0014] This invention also provides a map data acquisition method. Based on the received data collection task strategy, confirm that the current time is within the data collection task period; Confirm that the relationship between the current vehicle latitude and longitude and the start and end points of the data collection task strategy meets the preset conditions; Store point cloud data and image data for data acquisition; The collected point cloud data and image data are sliced ​​at predetermined unit time intervals and uploaded in slice mode.

[0015] According to one embodiment of the present invention, confirming that the relationship between the current vehicle latitude and longitude and the start and end points of the data collection task strategy meets preset conditions includes: It is confirmed that the straight-line distance between the current vehicle latitude and longitude and the starting point, and the straight-line distance between the current vehicle latitude and longitude and the ending point are both less than the straight-line distance between the starting point and the ending point.

[0016] This invention significantly optimizes the efficiency, cost, and accuracy of high-precision map data collection by reusing logistics operation vehicles as data collection carriers and combining vehicle-cloud collaboration mechanisms with refined data processing strategies. This invention eliminates the need for dedicated data collection vehicles. It filters logistics operation vehicles that frequently pass through target areas in historical trajectories through a data collection vehicle generation module, directly reusing their existing LiDAR, cameras, and other equipment to complete data collection. This reduces the costs of purchasing, maintaining, and scheduling dedicated vehicles and solves the resource waste problem of "hoarding a large number of data collection vehicles" in the traditional model.

[0017] In addition, the vehicle-side data acquisition module adopts a "pre-defined unit time slice upload" mechanism, which slices the collected data in real time and uploads it to the cloud, avoiding the delay caused by traditional batch backhaul. At the same time, the data acquisition task distribution module generates multiple tasks by "interrupting the trajectory at the first predetermined distance", which realizes the precise division and dynamic distribution of the data acquisition area. The cycle from data acquisition to data upload is significantly shortened compared with the traditional mode, improving the freshness of map data.

[0018] Furthermore, the vehicle-side data acquisition module ensures accurate location of the data collection area (within 30 meters) even at high speeds by determining that "the straight-line distance between the current latitude and longitude and the starting and ending points is less than the distance between the starting and ending points," combined with a position determination frequency of once per second. This reduces data redundancy or omissions. Simultaneously, the data storage module provides a reliable data foundation for area positioning by specifically storing the unique ID of the electronic fence and the vehicle's historical trajectory. Finally, the vehicle generation module performs trajectory clustering based on heading angle (e.g., 30 degrees) and spatial distance (e.g., 20 meters) thresholds to select the optimal data collection vehicle, ensuring a match between the task and the vehicle's driving patterns. The task tracking module monitors the data upload status in real time through vehicle coding and task cycles, achieving full-process traceability and improving the controllability and dynamic response capability of the data collection task.

[0019] This invention effectively solves the core pain points of traditional map data collection, such as high cost, long cycle, and inaccurate positioning, through an integrated solution of "reusable carrier - precise positioning - real-time upload - dynamic collaboration". It is especially suitable for the rapid update needs of high-precision maps in logistics scenarios. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0021] Figure 1 This is a block diagram of the map data acquisition system of the present invention.

[0022] Figure 2 This is a block diagram of the vehicle generation module.

[0023] Figure 3 This is a block diagram of the task distribution module.

[0024] Figure 4 This is a block diagram of the vehicle-side data acquisition module.

[0025] Figure 5 This is a block diagram of the task tracking module.

[0026] Figure 6 This is a flowchart of the trajectory clustering module.

[0027] Figure 7 This is a flowchart of the relationship judgment module.

[0028] Figure 8This is a flowchart of the map data collection method.

[0029] Explanation of reference numerals in the attached figures: 100. Map Data Acquisition System; 101. Vehicle Generation Module; 102. Task Distribution Module; 103. Vehicle-side Acquisition Module; 104. Task Tracking Module; 105. Data Storage Module; 1011. Data Acquisition Module; 1012. Trajectory Clustering Module; 1013. Task Input Data Determination Module; 1021. Task Input Module; 1022. Task Generation Module; 1023. Distribution Module; 1031. Task Receiving Module; 1032. Relationship Judgment Module; 1033. Data Acquisition Module; 1034. Data Upload Module; 1035. Latitude and Longitude Acquisition Module; 1041. Data Download Module; 1042. Map Production Module. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0031] The following is combined Figures 1-8 This invention describes a map data acquisition system and method according to embodiments of the present invention.

[0032] Figure 1 This is a block diagram of the map data acquisition system provided in an embodiment of the present invention. Figure 1 As shown, the system includes: The vehicle generation module 101 is used to acquire vehicle and trajectory data within the electronic fence of the collection area during historical periods, perform trajectory clustering based on thresholds, and determine the input data for the collection task.

[0033] Specifically, such as Figure 2As shown, the vehicle generation module 101 includes: a data acquisition module 1011, a trajectory clustering module 1012, and a data input determination module 1013. The workflow includes: First, the data acquisition module 1011 receives the unique ID of the electronic fence corresponding to the target collection area submitted by the operator. Using this as a spatial range benchmark, it retrieves the driving trajectory data of all vehicles passing through the electronic fence within a preset time period (e.g., 10 days) from the data storage module. Next, the trajectory clustering module 1012 performs cluster analysis on multiple trajectories for each vehicle, counting the number of trajectories in each cluster to reflect the vehicle's travel patterns in the area. Finally, the data input determination module 1013 selects the vehicles with the highest frequency of occurrence in the target area from historical trajectories and their corresponding trajectory data, using this as the basic input data for generating subsequent collection tasks, ensuring that the selected vehicles efficiently match the collection requirements and reducing resource waste.

[0034] It should be noted that the data collection vehicles used in this invention are preferably logistics operation vehicles. Because logistics operation vehicles have fixed routes and regular departure frequencies, they can cover specific transportation channels frequently, which highly matches the area coverage requirements for map data collection. At the same time, these vehicles are usually equipped with LiDAR, high-definition cameras, and high-precision positioning equipment (for their own navigation and safe operation), eliminating the need for additional dedicated data collection hardware. Existing equipment can be directly reused to complete data collection, significantly reducing hardware investment costs. Furthermore, the operating routes of logistics vehicles mostly revolve around core transportation networks such as logistics hubs and main roads, which are precisely the scenarios where the need for high-precision map data updates is most urgent. By reusing logistics vehicles, data collection and transportation tasks can be carried out in a coordinated manner, avoiding interference with the original vehicle operation plan and improving the efficiency and freshness of map data collection without increasing additional manpower costs.

[0035] The data collection task distribution module 102 is used to generate a data collection task strategy based on the data collection task input data and distribute it to the vehicle terminal.

[0036] Specifically, such as Figure 3As shown, the data acquisition task distribution module includes: a data acquisition task input module 1021, a data acquisition task strategy generation module 1022, and a distribution module 1023. The workflow of the data acquisition task distribution module 102 includes: First, the data acquisition task input module 1021 receives the data acquisition task cycle set by the operator, such as a specific date range or time period, and, based on the vehicle code and corresponding trajectory data determined by the vehicle generation module, divides the complete trajectory into multiple continuous segment trajectories at a first predetermined distance, wherein the first predetermined distance is preferably 500 meters; for each segment trajectory, the first trajectory point is extracted as the starting point and the last trajectory point is extracted as the ending point, thereby defining the spatial range of each segment acquisition area; then, the data acquisition task strategy generation module 1022 integrates the vehicle code, task cycle, and the starting and ending point information of all segment trajectories to generate a structured data acquisition task strategy containing "vehicle code + task cycle + multiple acquisition areas (starting point - ending point)"; finally, the distribution module 1023 distributes the strategy to the vehicle-side system of the corresponding vehicle through the vehicle-cloud collaborative channel, ensuring that the vehicle-side clearly knows when and on which road sections the data acquisition task is performed, thereby achieving accurate matching between the data acquisition task and the vehicle's driving plan.

[0037] In this way, the data collection task distribution module 102 divides the trajectory into fixed distances and clarifies the start and end points of each data collection area. Combined with the task cycle, it forms a precise structured strategy, which not only ensures that the data collection task is highly consistent with the vehicle's driving route and avoids invalid data collection, but also achieves fine coverage of complex routes through multi-segment area division. At the same time, relying on the efficient distribution strategy of the vehicle-cloud collaborative channel, it ensures that the vehicle receives and executes the task in a timely manner, which greatly improves the targeting and execution efficiency of the task distribution, laying the foundation for shortening the data collection cycle and improving the freshness of map data.

[0038] The vehicle-side data acquisition module 103 is used to acquire data based on the relationship between the current vehicle-side latitude and longitude, start point, and end point within the data acquisition task cycle, according to the data acquisition task strategy.

[0039] Specifically, such as Figure 4As shown, the vehicle-side data acquisition module 103 includes: a data acquisition task strategy receiving module 1031, a relationship judgment module 1032, a data acquisition module 1033, a data upload module 1034, and a latitude and longitude acquisition module 1035. The data acquisition task strategy receiving module 1031 first receives and parses the strategy information transmitted by the data acquisition task issuing module 102, including the vehicle code, task cycle, and the coordinates of the start and end points of each acquisition segment. During the vehicle's movement, the relationship judgment module 1032 monitors in real time whether the current time is within the task cycle. If not, data acquisition is not initiated. If it is within the cycle, the latitude and longitude acquisition module 1035 continuously acquires the current latitude and longitude of the vehicle. The relationship judgment module 1032 uses a spatial distance algorithm to determine the relationship between the current position and the start and end points of the corresponding acquisition area, determines that the vehicle has entered the acquisition segment, and automatically starts devices such as LiDAR and cameras. The data acquisition module 1033 begins to store point cloud data and image data. When the vehicle reaches the end of the segment and the above distance conditions are no longer met, the data acquisition module 1033 automatically stops the acquisition for that segment, ensuring that data acquisition is only performed within the target area and reducing the generation of invalid data. Finally, the data upload module 1034 uploads the acquired data to the cloud.

[0040] In this way, the vehicle-mounted data acquisition module 103 achieves fully automatic triggering and stopping of the acquisition process through a dual mechanism of time period verification and spatial distance judgment. It can accurately lock the target acquisition area without manual intervention, which not only avoids missed or incorrect acquisition of areas due to delays caused by manual operation, but also reduces the occupation of storage and transmission resources by invalid data. At the same time, the dynamic response mode based on preset conditions can adapt to the trajectory fluctuations during vehicle driving, and can maintain accurate coverage of the acquisition area even in complex road conditions. Combined with real-time stored point cloud and image data, it provides high-quality raw materials for the efficient production of subsequent map data, further shortening the entire process cycle from acquisition to data application.

[0041] The task tracking module 104 is used to query the data uploaded by the vehicle based on the vehicle code and the collection task cycle.

[0042] Specifically, such as Figure 5As shown, the task tracking module 104 includes a data download module 1041 and a map production module 1042. The workflow of the task tracking module 104 includes: the operator inputs the target vehicle's code and the corresponding data collection task cycle into the system; the data download module 1041 retrieves all data uploaded by the vehicle within the specified cycle from the data storage module based on these two key parameters, including point cloud data, image data, and corresponding timestamps and location information for each collection area; simultaneously, the data download module 1041 verifies the integrity of the data, calculates the matching degree between the uploaded data and the collection areas specified in the task strategy, and generates a tracking report containing information such as "completed collection sections / incomplete collection sections," "data upload time," and "data size"; the operator can monitor the task execution progress in real time through this report. If data loss or upload anomalies are detected, a timely supplementary collection mechanism can be triggered, ensuring that the entire data collection task process is traceable and controllable, further guaranteeing the integrity and timeliness of map data collection. If there are no abnormalities in the data, the download data module 1041 marks the complete collection data within the period as "available" and pushes it to the production map module 1042 for subsequent map data stitching, labeling, and updating. At the same time, the production map module 1042 automatically records the completion status of this task, including key indicators such as the total number of road segments collected, the actual number completed, and the total data volume, and forms a task archive report to provide a reference for the planning of similar collection tasks in the future.

[0043] In this way, the task tracking module establishes a two-dimensional retrieval mechanism of "vehicle code + task cycle" to achieve accurate traceability and progress monitoring of collected data. This ensures the integrity and validity of data uploads and provides a basis for timely handling of anomalies, avoiding map update delays caused by missing data. At the same time, by generating standardized task reports, it not only achieves closed-loop management of the collection process but also provides data support for subsequent task optimization, further improving the controllability and overall efficiency of map data collection and helping to form an efficient collaborative link of "collection-tracking-production".

[0044] According to an embodiment of the present invention, a map data acquisition system 100 is provided, wherein the trajectory clustering module 1012 of the vehicle generation module 101 is used to confirm that multiple trajectories of the vehicle are similar when the heading angle of the vehicle is less than or equal to a first angle threshold and the spatial distance of the vehicle is less than or equal to a first length threshold, and to group the similar trajectories into the same cluster.

[0045] Specifically, such as Figure 6As shown, in step S601, the trajectory clustering module 1012 of the vehicle generation module 101 analyzes multiple driving trajectories of the selected vehicles during the historical period segment by segment, calculating the heading angle of each sampling point in each trajectory and the spatial distance between trajectories. When the heading angle deviation of corresponding sampling points in two trajectories is less than or equal to the first angle threshold (preferably 30 degrees) in step S602, and the overall spatial distance of the trajectories (calculated by the Hausdorff distance algorithm) is less than or equal to the first length threshold (preferably 20 meters) in step S603, it is determined in step S604 that these two trajectories are highly similar and belong to the same trajectory cluster under the same driving pattern. By performing this type of clustering processing on all historical trajectories, trajectories with similar characteristics are aggregated into the same cluster, thereby reflecting the stable driving path pattern of vehicles in the target area, providing data support for subsequent screening of high-frequency and high-matching vehicles, ensuring the fit between the driving trajectory of the selected vehicle and the collection area, reducing invalid collection paths, and improving task execution efficiency.

[0046] In this way, the vehicle generation module 101 achieves trajectory clustering through dual threshold verification of heading angle and spatial distance, which can accurately identify the stable driving pattern of vehicles in the target area and ensure that the clustered trajectory clusters have a high degree of consistency. This clustering method based on historical trajectory patterns not only provides a scientific basis for selecting vehicles that frequently pass through the collection area, but also reduces the misselection of vehicles due to accidental deviation from the route, making the driving trajectory of the selected collection vehicles more closely match the target area. This minimizes the proportion of invalid paths in the subsequent collection process, improves the targeting and overall efficiency of map data collection, and lays the foundation for achieving low-cost, high-precision map data collection.

[0047] According to an embodiment of the present invention, a map data acquisition system 100 is provided, wherein the vehicle generation module 101 is used to acquire the vehicle and trajectory data that appear most frequently during the historical period as input data for the acquisition task.

[0048] Specifically, after the vehicle generation module 101 completes trajectory clustering, the data input module 1013 statistically analyzes the frequency of vehicles corresponding to each cluster passing through the target data collection area within a historical period (preferably 10 days), including indicators such as the number of times vehicles enter the electronic fence, cumulative dwell time, and trajectory coverage density. Subsequently, based on these indicators, the vehicles are sorted, and the vehicles with the highest frequency of occurrence and optimal trajectory stability in the target area are selected. Their corresponding trajectory data (including typical paths after clustering, travel times, etc.) are then determined as the data input for the data collection task. This process ensures that the selected vehicles best match the geographical scope and traffic needs of the data collection area. Their inherent high-frequency driving patterns reduce additional scheduling costs, while ensuring that the collected data covers the core road sections of the target area, providing highly adaptable basic data for subsequent task deployment, further improving the efficiency and accuracy of the entire data collection process.

[0049] In this way, the vehicle generation module 101, by focusing on frequently passing vehicles, can fully utilize their stable driving trajectories and regular operating cycles, ensuring that the data collection task highly overlaps with the vehicles' inherent routes, thus reducing unnecessary mileage and resource waste from the source. Simultaneously, the data collected from frequently appearing vehicles can more comprehensively cover the traffic scenarios of the target area, especially showing higher sensitivity to road condition changes in core areas such as logistics hubs and main roads. The generated data is also more in line with actual operational needs, providing more representative raw materials for subsequent map data updates, further enhancing the freshness and application value of the map data. Furthermore, based on the selection logic of high-frequency vehicles, reliance on vehicle dispatching can be reduced, lowering the cost increases caused by temporary vehicle deployment, making the entire data collection process more aligned with the actual operational characteristics of logistics scenarios, achieving a dual optimization of efficiency and economy.

[0050] According to an embodiment of the present invention, a map data acquisition system 100 includes an acquisition task distribution module 102 configured to: based on the vehicle code and trajectory data generated by the vehicle generation module 101, interrupt the trajectory at a first predetermined distance to generate multiple segment trajectories, and extract a start point and an end point within each segment trajectory, wherein the start point is the first trajectory point of the segment trajectory, and the end point is the last trajectory point of the segment trajectory; the acquisition task distribution module 102 is configured to generate an acquisition task strategy based on the vehicle code, the acquisition task cycle, and multiple acquisition areas composed of the start point and the end point.

[0051] Specifically, the task input module 1021 of the task distribution module 102 first obtains the filtered vehicle codes and their corresponding complete trajectory data from the vehicle generation module 101. Then, the task strategy generation module 1022 divides the trajectory into segments according to a first predetermined distance (preferably 500 meters), splitting the long-distance trajectory into multiple continuous segment trajectories. For each segment trajectory, the latitude and longitude of its first trajectory point are precisely extracted as the starting point, and the latitude and longitude of its last trajectory point are extracted as the ending point, thus forming a segmented collection area unit of "starting point - ending point". Subsequently, the task strategy generation module 1022 integrates the vehicle codes, the collection task cycle set by the operator (such as from a certain date to a certain day), and the start and end point information of all segmented collection areas to generate a structured collection task strategy. The strategy includes key information in a clear format, such as "the time period during which the vehicle needs to perform data collection" and "the specific road segments to be covered in each time period (defined by the start and end points)". Finally, the distribution module 1023 distributes the information to the vehicle's onboard terminal through the vehicle-cloud collaborative network, enabling the vehicle to clearly identify the task boundaries and execution requirements, and providing accurate instructions for subsequent automatic data collection.

[0052] In this way, the task distribution module 102 achieves a refined breakdown of complex trajectories, ensuring that the range of each collection segment is clear and controllable. This not only adapts to the continuity of routes in actual vehicle driving, but also reduces the pressure of single-segment data collection through segmented management. At the same time, by deeply binding vehicle identifiers with time and spatial parameters, the task strategy has a unique orientation, avoiding the problems of multi-vehicle task confusion or overlapping area collection, further improving the efficiency of vehicle-cloud collaboration and the executability of collection tasks.

[0053] According to an embodiment of the present invention, a map data acquisition system 100 is provided, wherein the vehicle-mounted acquisition module 103 is used to: confirm that the current time is within the acquisition task cycle, confirm that the relationship between the current vehicle-mounted latitude and longitude and the starting point and the ending point meets preset conditions, store point cloud data and image data for data acquisition, and upload the acquired point cloud data and image data.

[0054] Specifically, such as Figure 7As shown, after receiving the acquisition task strategy, the relationship judgment module 1032 of the vehicle-mounted acquisition module 103 compares the local time with the task cycle in real time during step S701. Subsequent acquisition logic is only initiated when the current time falls within the acquisition period specified by the strategy (e.g., 8:00-18:00 daily), avoiding unnecessary energy consumption during non-task periods. During the acquisition period, the latitude and longitude acquisition module 1035 of the data acquisition module 1033 acquires the current latitude and longitude once per second through the vehicle-mounted positioning system (e.g., GNSS). In step S702, the relationship judgment module 1032 continuously calculates the straight-line distance to the start and end points of the current segment to be acquired using a spatial distance algorithm. When the preset conditions are met, the relationship judgment module 1032 automatically activates the lidar and high-definition camera in step S703, and the data acquisition module 1033 starts to collect point cloud data and image data at a preset frame rate (e.g., lidar 10Hz, camera 30 frames / second), and stores the two types of data in the vehicle's local cache (association methods include timestamp alignment and location information binding).

[0055] When the vehicle reaches the end of a segment, causing the distance condition to be unmet, the data acquisition module 1033 immediately stops acquiring data for that segment and simultaneously triggers the data upload module 1034 to upload the data. If a network interruption occurs, the data upload module 1034 will temporarily store the data and automatically resume uploading once the network is restored, ensuring data integrity.

[0056] Through this fully automated mechanism of "time verification - location judgment - real-time storage - upload", the vehicle-side data acquisition module 103 can strictly limit the collection range and time period, and realize data collection and transmission at the same time, which greatly reduces the cost of manual intervention and the risk of data loss. At the same time, it ensures the temporal and spatial correlation of uploaded data, providing structured raw materials for cloud data processing.

[0057] According to an embodiment of the present invention, a map data acquisition system 100 is provided, wherein the vehicle-mounted acquisition module 103 is used to confirm that the straight-line distance between the current vehicle latitude and longitude and the starting point and the straight-line distance between the current vehicle latitude and longitude and the ending point are both less than the straight-line distance between the starting point and the ending point, and to confirm that a preset condition is met.

[0058] Specifically, after the latitude and longitude acquisition module 1035 of the vehicle-side acquisition module 103 obtains the current latitude and longitude coordinates in real time through the vehicle positioning system, the relationship judgment module 1032 calls the Haversine formula (spherical distance calculation formula) to calculate two key distances: one is the straight-line distance between the current position and the starting point of the segment trajectory (denoted as D1), and the other is the straight-line distance between the current position and the ending point of the segment trajectory (denoted as D2). At the same time, the pre-generated acquisition task strategy already includes the straight-line distance between the starting point and the ending point of the segment (denoted as D_total). The module compares D1 and D2 with D_total respectively, and only when both conditions "D1 < D_total" and "D2 < D_total" are met simultaneously is it determined that the vehicle is in the acquisition area corresponding to the segment trajectory.

[0059] This judgment logic essentially defines the collection range through geometric relationships: using the line segment formed by the starting and ending points as the reference, the current position must be located within the extension range on both sides of this line segment (rather than outside the line segment's extension line), thus ensuring that the vehicle travels within the reasonable coverage area of ​​the segment trajectory. For example, when the distance between the starting and ending points of a segment trajectory is 500 meters (D_total = 500 meters), if the current position is 300 meters from the starting point (D1 = 300 meters) and 250 meters from the ending point (D2 = 250 meters), and both are less than 500 meters, then the collection condition is met; if the vehicle travels outside the end point of the segment, resulting in D1 = 600 meters (greater than D_total), then it is automatically determined to have left the collection area.

[0060] Through this dynamic distance verification mechanism, the vehicle-side data acquisition module 103 can maintain a judgment frequency of once per second even when the vehicle is traveling at high speed (such as 80 km / h), keeping the positioning error of the acquisition area within 30 meters. This not only avoids missed data collection due to slight deviation of the vehicle from the trajectory, but also prevents invalid data collection beyond the target road segment, significantly improving the accuracy and validity of the acquisition area.

[0061] According to an embodiment of the present invention, a map data acquisition system 100 is provided, wherein the vehicle-mounted acquisition module 103 is used to slice the acquired point cloud data and image data at predetermined unit time intervals and upload the acquired point cloud data and image data in slice mode.

[0062] Specifically, when storing point cloud data and image data, the data acquisition module 1033 of the vehicle-mounted acquisition module 103 activates a built-in time slicing mechanism to segment the real-time acquired data stream according to a predetermined unit time (e.g., 5 minutes). After each time slice is completed, the data acquisition module 1033 automatically adds metadata tags to the slice data, including the timestamp corresponding to the slice (accurate to the second), the average latitude and longitude during the acquisition period, and the start / end point identifiers of the corresponding segment trajectory, ensuring that each slice data can be accurately associated with the specific acquisition time and spatial location.

[0063] During the upload phase, the data acquisition module 1034 prioritizes pushing sliced ​​data to the cloud data storage module in real time via a vehicle-cloud collaborative network (such as 5G). If a weak or interrupted network signal occurs, the sliced ​​data will first be cached on the vehicle's local storage (such as a solid-state drive), and will automatically resume uploading in chronological order after the network is restored to avoid data loss. After receiving the sliced ​​data, the cloud can quickly complete data stitching and spatiotemporal alignment based on the timestamp and location information in the metadata, and can start the initial processing flow without waiting for all data to be collected.

[0064] Compared to traditional batch upload methods, this slice upload mode significantly reduces the data volume of a single transmission (e.g., the data volume of a 5-minute slice is about 1 / 10 of that of traditional batch transmission), reduces network bandwidth consumption and transmission latency, and shortens the cycle from map data collection to cloud availability to minutes. At the same time, through time-divisional management, it is easier to extract and analyze road condition data for specific time periods, improving the flexibility and efficiency of data processing and providing real-time support for the dynamic updating of high-precision maps.

[0065] According to an embodiment of the present invention, a map data acquisition system 100 further includes a data storage module 105, which is used to store the vehicle code, driving date and trajectory data of each vehicle; and to store the electronic fences of all areas to be collected, the corresponding unique IDs, and to receive and store the data collected by the vehicles.

[0066] Specifically, the data storage module 105 constructs a multi-dimensional data storage architecture: For basic vehicle information, it is stored in a hierarchical structure of "vehicle code - driving date - trajectory data". The trajectory data includes dynamic parameters such as latitude and longitude, heading angle, and driving speed once per second, and supports quick retrieval by vehicle, date, or road segment; For collection area information, it adopts an association mode of "electronic fence geometric parameters - unique ID - associated road segment attributes". The electronic fence is stored in the form of polygon coordinate strings, and the unique ID corresponds one-to-one with the area identifier in the collection task strategy, which facilitates quick location of the target collection range.

[0067] When receiving data uploaded from the vehicle, the data storage module 105 automatically classifies and stores the data slices based on their metadata (such as vehicle code, timestamp, and collection area ID). Point cloud data and image data are stored separately in their respective structured databases, and an association index is established between the two using timestamps and location information to ensure subsequent joint queries along the "time-space" dimension. Simultaneously, the data storage module 105 has a data verification function, performing integrity and format verification on the uploaded slice data. If data corruption or missing data is detected, it will report the anomaly to the task tracking module, triggering a re-collection mechanism.

[0068] It should be noted that, in this invention, the data storage module is preferably located in the cloud. Using cloud storage overcomes the capacity limitations of local vehicle storage, supports long-term archiving of massive historical trajectory data and collected data, and leverages the distributed computing capabilities of the cloud to quickly respond to multi-dimensional data query requests. This provides efficient data support for trajectory clustering in the vehicle generation module, progress monitoring in the task tracking module, and subsequent map data production, achieving seamless data flow throughout the entire process of "collection-storage-application".

[0069] Using the cloud as the data storage module fully leverages the technological advantages of vehicle-cloud collaboration: On one hand, by establishing a real-time data interaction channel with the vehicle, the cloud can immediately trigger data verification and preprocessing processes after receiving the tiled data uploaded by the vehicle, achieving seamless integration of "vehicle-side data collection - cloud storage - data production," significantly shortening the turnaround time of map data from collection to application, and greatly improving the dynamic update speed of high-precision maps; on the other hand, with its powerful computing and storage capabilities, the cloud can simultaneously connect to the data collection data of multiple logistics vehicles, and integrate the collected information from different vehicles and time periods through a unified data standard, eliminating data silos and ensuring the global consistency of collected data; in addition, the vehicle-cloud collaboration mode supports the cloud in dynamically optimizing the collection task strategy based on historical stored data. For example, by analyzing the historical trajectory density of a certain area, the granularity of subsequent collection task segmentation can be automatically adjusted, enabling collection resources to be more accurately deployed to road sections with insufficient data freshness, and realizing intelligent scheduling of collection tasks. This cloud-based vehicle-cloud collaboration mechanism not only enhances the reliability and scalability of data storage, but also promotes the formation of a closed loop of "perception-decision-execution-feedback" in the entire map data acquisition system through real-time data flow and deep linkage, significantly improving the intelligence level and practical application value of high-precision map data acquisition.

[0070] On the other hand, the present invention also provides a map data acquisition method, comprising: confirming that the current time is within the acquisition task cycle based on the received acquisition task strategy; confirming that the relationship between the current vehicle latitude and longitude and the start point and end point of the acquisition task strategy meets preset conditions; storing point cloud data and image data for data acquisition; slicing the acquired point cloud data and image data into slices at predetermined unit time intervals, and uploading the acquired point cloud data and image data in slice mode.

[0071] Specifically, such as Figure 8 As shown, the execution flow of this map data collection method is as follows: After receiving the collection task strategy issued by the cloud in step S801, the vehicle first parses the vehicle code, task period (such as 9:00-18:00 daily from August 1 to August 7, 2025) and the coordinates of the start and end points of each collection area in steps S802 and S803 to ensure that the local task is consistent with the cloud instructions.

[0072] During vehicle operation, the vehicle-side system synchronizes the current time every second. In step S802, this time is compared with the task cycle. Subsequent data acquisition logic is only initiated if the time period is within the specified time frame. If the cycle is exceeded, the acquisition equipment is automatically shut down to save energy. When the acquisition period is in progress, in step S803, the system obtains the current latitude and longitude in real time through a high-precision positioning module (such as GNSS+IMU combined positioning). It continuously calculates the straight-line distance to the starting and ending points of the current segment to be acquired using a spatial distance algorithm. The vehicle is determined to have entered the effective acquisition area only when both the conditions of "distance from the current position to the starting point < distance between the starting and ending points" and "distance from the current position to the ending point < distance between the starting and ending points" are met. Then, in step S804, the LiDAR (such as a 128-line LiDAR) and high-definition camera (such as an 8-megapixel camera) are activated. Point cloud data and image data are acquired according to preset parameters (LiDAR frame rate 10Hz, camera 30 frames / second). The two types of data are associated and stored in the vehicle-side local cache (such as a 1TB solid-state drive) through timestamps and location information.

[0073] During data acquisition, in step S805, the system slices the cached data at predetermined time intervals (e.g., 5 minutes). For each slice completed, metadata is generated, including the slice ID, timestamp range (e.g., 8:00:00-8:05:00), average latitude and longitude, and the corresponding acquisition segment ID. The sliced ​​data (including metadata) is then uploaded to the cloud data storage module via the 5G vehicle-to-cloud communication module. In the event of a network interruption, the system temporarily stores the unuploaded slices locally and automatically resumes transmission in chronological order once the network is restored, ensuring data continuity.

[0074] After receiving the tile data in the cloud, it performs classification, storage, and verification based on metadata. The vehicle-side then deletes the corresponding tile from its local cache after successful upload, freeing up storage resources. This process automates the entire workflow from task reception, spatiotemporal verification, data collection to tile upload, ensuring both the spatiotemporal accuracy of the collected data and improving data transmission efficiency through the tile upload mechanism. This provides a reliable methodological support for the rapid updating of high-precision maps.

[0075] This map data acquisition method, through a complete process design of "time verification - spatial positioning - data storage - tile upload," demonstrates significant technical effects: In the time dimension, by confirming that the current time is within the acquisition task cycle, it avoids ineffective energy consumption and data redundancy during non-task periods, ensuring the synchronization of acquisition activities with the planned time. In the spatial dimension, by determining the acquisition area based on the distance relationship between the current vehicle's latitude and longitude and the starting and ending points, it can accurately locate the target road segment, controlling the positioning error within 30 meters and reducing missed or incorrect acquisitions caused by trajectory deviations. In terms of data processing, the associated storage of point cloud and image data ensures the spatiotemporal consistency of multi-source data, while the predetermined unit time tile upload mechanism realizes "acquiring and transmitting simultaneously," reducing the bandwidth pressure of a single transmission, enabling the cloud to receive and process data in real time, shortening the map update cycle from the traditional days to hours, and significantly improving the freshness of high-precision maps. Overall, this method reduces manual intervention while taking into account acquisition efficiency, data accuracy, and real-time performance, perfectly adapting to the needs of rapid updates of high-precision maps in logistics scenarios, effectively reducing acquisition costs and enhancing the value of data applications.

[0076] According to an embodiment of the present invention, a map data acquisition method is provided in which the step of confirming that the relationship between the current vehicle latitude and longitude and the starting point and the ending point of the acquisition task strategy meets the preset conditions includes: confirming that the straight-line distance between the current vehicle latitude and longitude and the starting point and the straight-line distance between the current vehicle latitude and longitude and the ending point are both less than the straight-line distance between the starting point and the ending point.

[0077] Specifically, the vehicle-side system obtains the starting point (S), ending point (E), and distance (L) between the two points from the task strategy. It then calculates the distance (L1) from the current position (P) to S and the distance (L2) to E in real time. Only when L1 < L and L2 < L is the system considered to be within the valid data collection area. This precise targeting of the collection range avoids missed or invalid data collection, ensuring data quality. The logic is similar to the mechanism in the vehicle-side data collection module that determines the collection area based on distance relationships: using the line segment distance between the starting and ending points as a benchmark, the system calculates the distance between the current position and the two points in real time and compares it with the benchmark value to form a dynamic verification of the spatial range. This logic maintains positioning accuracy within 30 meters even in high-speed driving scenarios, avoiding missed data collection due to trajectory deviations and preventing invalid data collection beyond the target road segment. This is consistent with the corresponding technical implementation principle in the system and will not be elaborated further here.

[0078] This technical solution achieves dynamic and precise positioning of the data collection area through a "dual distance less than the baseline distance" judgment logic, exhibiting significant technical advantages: Firstly, its spatial verification mechanism based on geometric relationships maintains positioning accuracy within 30 meters even in high-speed vehicle scenarios (e.g., 80 km / h), effectively avoiding missed data collection due to lane departures, turns, and other scenarios. Simultaneously, it strictly limits the collection range, avoiding invalid data collection on non-target road sections and reducing redundancy in subsequent data processing. Secondly, this logic does not rely on fixed markers (such as road signs in traditional solutions), dynamically adapting to trajectory fluctuations solely through latitude and longitude distance calculations. This makes it more adaptable to complex road conditions, particularly suitable for scenarios such as urban arterial roads and logistics hubs with high-frequency logistics vehicle traffic, ensuring that the collected data accurately covers the core road sections of the target area. Overall, this judgment method improves the spatial accuracy of the collected data while reducing reliance on environmental markers, providing high-quality raw material support for the efficient production of map data.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A map data acquisition system, characterized in that, The system includes: The vehicle generation module is used to acquire vehicle and trajectory data within the electronic fence of the collection area during historical periods, perform trajectory clustering based on thresholds, and determine the input data for the collection task. The data collection task distribution module is used to generate a data collection task strategy based on the data input for the data collection task and distribute it to the vehicle terminal. The vehicle-side data acquisition module is used to acquire data based on the relationship between the current vehicle-side latitude and longitude, start point, and end point within the data acquisition task cycle, according to the data acquisition task strategy. The task tracking module is used to query vehicle-uploaded data based on vehicle codes and data collection task cycles.

2. The map data acquisition system according to claim 1, characterized in that, The vehicle generation module is used to determine if multiple trajectories of a vehicle are similar when the vehicle's heading angle is less than or equal to a first angle threshold and the vehicle's spatial distance is less than or equal to a first length threshold, and to group the similar trajectories into the same cluster.

3. The map data acquisition system according to claim 2, characterized in that, The vehicle generation module is used to obtain the vehicle and trajectory data that appear most frequently during the historical period as input data for the data collection task.

4. The map data acquisition system according to claim 1, characterized in that, The task distribution module is used for: Based on the vehicle code and trajectory data generated by the vehicle generation module, the trajectory is interrupted at a first predetermined distance to generate multiple segment trajectories. A start point and an end point are taken in each segment trajectory, wherein the start point is the first trajectory point of the segment trajectory and the end point is the last trajectory point of the segment trajectory. The data acquisition task distribution module is used to generate a data acquisition task strategy based on the vehicle code, the data acquisition task cycle, and multiple data acquisition areas consisting of the start point and the end point.

5. The map data acquisition system according to claim 1, characterized in that, The vehicle-mounted data acquisition module is used for: Confirm that the current time is within the data collection task cycle, and confirm that the relationship between the current vehicle latitude and longitude and the starting point and the ending point meets preset conditions. Store point cloud data and image data for data acquisition. Upload the collected point cloud data and image data.

6. The map data acquisition system according to claim 5, characterized in that, The vehicle-mounted data acquisition module is used to confirm that the straight-line distance between the current vehicle-mounted latitude and longitude and the starting point and the straight-line distance between the current vehicle-mounted latitude and longitude and the ending point are both less than the straight-line distance between the starting point and the ending point, thus confirming that the preset conditions are met.

7. The map data acquisition system according to claim 5, characterized in that, The vehicle-mounted acquisition module is used to slice the acquired point cloud data and image data at predetermined unit time intervals and upload the acquired point cloud data and image data in slice mode.

8. The map data acquisition system according to claim 1, characterized in that, It also includes a data storage module for storing the vehicle code, driving date, and trajectory data of each of the vehicles; It is also used to store the electronic fences of all areas that need to be collected, their corresponding unique IDs, and to receive and store the data collected by the vehicle.

9. A map data acquisition method, characterized in that, Based on the received data collection task strategy, confirm that the current time is within the data collection task period; Confirm that the relationship between the current vehicle latitude and longitude and the start and end points of the data collection task strategy meets the preset conditions; Store point cloud data and image data for data acquisition; The collected point cloud data and image data are sliced ​​at predetermined unit time intervals and uploaded in slice mode.

10. The map data acquisition method according to claim 9, characterized in that, The confirmation that the relationship between the current vehicle latitude and longitude and the start and end points of the data collection task strategy meets preset conditions includes: It is confirmed that the straight-line distance between the current vehicle latitude and longitude and the starting point, and the straight-line distance between the current vehicle latitude and longitude and the ending point are both less than the straight-line distance between the starting point and the ending point.

Images