Vehicle driving area identification method and system

Abstract

The invention provides a vehicle driving area identification method and system, and the method comprises the steps: building a second intersection node queue according to an SD map, the SD map comprises a first area, a third area, and a second area communicating the first area and the third area, and the second intersection node is an intersection where the second area is connected with the first area; when the vehicle is positioned to run to any node in the second intersection node queue, recording the position of the vehicle, and recording the position as a first position; after the vehicle passes through the first position, determining a current driving area of the vehicle according to the current position, the first position and the vehicle positioning information; and outputting the current driving area information. According to the vehicle driving area identification method, the problem of high-consumption and low-efficiency pain points in special scene identification in the prior art is solved, and the identification accuracy is improved.

Description

Technical Field

[0001] This application relates to the field of driving technology, and in particular to a method and system for identifying vehicle driving areas. Background Technology

[0002] With the rapid development of intelligent driving and autonomous driving technologies, various technologies enable vehicle positioning and rapid identification of driving areas. However, when vehicles travel in certain special areas, existing technologies lack sufficient accuracy in identifying these areas or require significant computational power. For example, in underground parking lots, existing SD maps only display surface road network information without elevation information, while underground parking information only includes the surface entrance / exit area, failing to distinguish between surface and underground information. When a vehicle enters these areas, if it cannot be identified as having entered, the in-vehicle display will indicate that the vehicle is still on the surface or issue a warning of deviation from the road. Existing methods for identifying entry and exit from underground parking lots rely on image recognition technology and deep learning networks, such as convolutional neural network object detection models. Typically, vehicle-mounted cameras collect a large amount of image data from underground parking lots (including vehicles, obstacles, signs, and structural features), and the model is trained using massive labeled datasets to learn and recognize patterns specific to these scenes. Although deep learning has made significant progress in image processing, existing methods have revealed a series of problems in practical applications, such as weak generalization ability, high data dependence, insufficient recognition accuracy, and high computational power consumption. Summary of the Invention

[0003] This application provides a method and system for identifying vehicle driving areas in order to solve the above-mentioned technical problems.

[0004] A method for identifying a vehicle's driving area includes: establishing a second intersection node queue based on an SD map, wherein the SD map includes a first region, a third region, and a second region connecting the first and third regions, and the second intersection node is an intersection connecting the second region to the first region; when a vehicle is located driving to any node in the second intersection node queue, the vehicle's position is recorded and designated as the first position; after the vehicle passes the first position, the current driving area of ​​the vehicle is determined based on the current position, the first position, and the vehicle's positioning information; and the current driving area information is output.

[0005] This solution utilizes SD maps and spatial topology logic to replace existing methods that rely on pixel-level image data analysis for vehicle driving area identification, addressing the high resource consumption and low efficiency of existing solutions in specific scenarios. Specifically, this solution abandons traditional image recognition techniques and deep learning networks for real-time analysis of high-resolution images. Instead, it uses predefined SD maps and intersection node topology data for localization. The amount of structured information such as node details and connectivity is only a fraction of that of image data, reducing the parallel computing burden on GPUs / TPUs. Through relevant node queues, vehicles quickly record their positions when a node is triggered, avoiding latency issues in image recognition and improving system real-time performance. By leveraging the structured data of SD maps, rather than relying on image features, the robustness of vehicle driving area identification is improved, and the dependence on massive amounts of labeled data is reduced. Furthermore, rapid localization and position recording enable real-time determination of the driving area, providing timely input for vehicle control (such as path planning or speed adjustment), thereby enhancing safety performance. Therefore, the vehicle driving area recognition method based on SD map in this technical solution improves the recognition accuracy of vehicles in special areas (such as underground parking lots) and provides a feasible path for vehicle driving area recognition technology in intelligent driving technology.

[0006] Furthermore, the step of establishing a second intersection node queue based on the SD map includes: detecting all links centered on the vehicle within a preset range on the SD map, as well as the start and end nodes of the links; establishing a directed graph based on all the links, as well as the start and end nodes of the links; traversing and searching for nodes located in the second region based on the directed graph to form a second node queue, and determining the second intersection node queue from the second node queue.

[0007] In this scheme, the second intersection node is used as a key geofence node connecting different areas, providing a static reference benchmark for vehicle location. Establishing a queue of second intersection nodes is the core foundation for improving the accuracy and efficiency of area identification.

[0008] Further, the step of searching for nodes located in the second region based on the directed graph traversal to form a second node queue includes: establishing and initializing a search queue based on the starting node of the link where the vehicle is located, with the starting node of the link where the vehicle is located being the head node of the search queue; sequentially extracting search nodes from the search queue, and determining the first ending node of all outgoing edges of each search node; if the first ending node belongs to a node in the second region and has not been extracted, adding the first ending node to the search queue as the search node, and adding the first ending node to the second node queue as the second node; completely traversing the search nodes and outputting the second node queue.

[0009] In this scheme, directed graph traversal search is a core method for vehicle driving area identification, used to model the road network. Nodes represent intersections or locations, and directed edges represent driving directions. Road networks naturally possess graph structure characteristics, and using a directed graph (with edge directions) can accurately represent road connectivity and driving restrictions. During the traversal process, after identifying the first region, a dedicated node queue (i.e., the second node queue) is maintained to process nodes in the second region. This can be implemented using a breadth-first search (BFS) algorithm. By separating queue management, the region identification efficiency is optimized, significantly improving computational efficiency and real-time performance.

[0010] Further, determining the second intersection node queue from the second node queue includes: sequentially extracting the second nodes from the second node queue, determining the second end node of all outgoing edges of each second node; if the second end node belongs to the second node queue, the second end node has a successor link, and the second end node has not been extracted, adding the second end node to the second intersection node queue as the second intersection node; completely traversing the second end node, and outputting the second intersection node queue.

[0011] In this scheme, a breadth-first search is used to traverse the directed graph, only filtering the intersection nodes connecting the second region to the first region and excluding irrelevant road nodes. This reduces the computational load for subsequent region judgments, accurately locates key nodes, and avoids redundant data interference.

[0012] Further, after the vehicle passes the first location, determining the vehicle's current driving area based on the current location, the first location, and the vehicle positioning information includes: if the height difference between the current location and the first location is greater than a first preset height threshold, the distance between the current location and the first location is less than a first preset distance threshold, and the vehicle signal strength does not exceed a preset signal strength threshold, then the vehicle is determined to have entered the third area; when the distance between the current location and the first location exceeds the first preset distance threshold, and the link where the current location is located belongs to the second area type, the current location is re-recorded as the first location.

[0013] In this solution, after the vehicle enters the second area, to accurately determine its current position during continued driving, a first position is used as a reference. The distance difference, height difference, and positioning signal characteristics between the current position and the first position are used to comprehensively determine whether the vehicle has entered a specific third area. This multi-dimensional data fusion judgment improves the accuracy and reliability of boundary determination compared to relying solely on traditional planar positioning. Height difference information provides a crucial third-dimensional criterion, effectively resolving ambiguities in the vertical direction. Distance difference provides a more accurate relative positional relationship than latitude and longitude alone. Positioning signal characteristics provide auxiliary evidence of entering a specific environmental type.

[0014] Furthermore, determining the current driving area of ​​the vehicle also includes: if the distance between the current position and the first position exceeds a first preset distance threshold, the link where the current position is located belongs to a first area type, and the vehicle signal strength exceeds a preset signal strength threshold and the duration exceeds a first time threshold, then the current driving area of ​​the vehicle is determined to be the first area.

[0015] This solution significantly improves the accuracy and scene adaptability of the first area identification by fusing multi-dimensional data such as distance difference, link type, signal strength and duration, making it particularly suitable for autonomous driving or intelligent navigation scenarios in complex road environments.

[0016] Furthermore, determining the vehicle's current driving area also includes: if the vehicle enters the third area, determining the link ID (linkid) of the link where the vehicle last entered the third area. rec The vehicle re-records its entry point into the third area as the first location; the third area includes multiple entrances and exits, and when the vehicle exits the third area, the link ID of the current location is determined. cur ; If linkid cur ≠linkid rec If the height difference between the current position and the first position is less than the second preset height threshold, and the link where the current position is located belongs to the second region type, then the current driving region of the vehicle is determined to be the second region, and the slot_flag flag of the vehicle's third region detection is set to true. If linkid cur =linkid rec If the height difference between the current position and the first position is less than the third preset height threshold, and the link where the current position is located belongs to the second region type, then the current driving area of ​​the vehicle is determined to be the second region, and the slot_flag flag of the vehicle's third region detection is set to true.

[0017] In this solution, when a vehicle exits from the third area, the linkid is used to determine whether the vehicle exited from the same entrance or exit. Then, based on different multi-dimensional data, the vehicle's driving area is determined, thereby improving the accuracy and scene adaptability of the second area recognition.

[0018] Furthermore, determining the vehicle's current driving area also includes: if the vehicle's third region detection success flag (slot_flag) is false, the link where the current position is located belongs to the first region type, the link where the first position is located belongs to the second region type, and the distance between the current position and the first position exceeds a second preset distance threshold, then the vehicle's current driving area is determined to be the first region; if the vehicle's third region detection success flag (slot_flag) is true, the link where the current position is located belongs to the first region type, and the link where the first position is located belongs to the second region type, then the vehicle's current driving area is determined to be the first region.

[0019] This solution employs a triple-judgment approach, utilizing a third-region detection success flag and a link type distance threshold, thus avoiding the risk of misjudgment from a single sensor or logic. The third-region detection success flag can mark specific scenarios, i.e., whether the vehicle has entered the third region. Based on this, combined with different link types and distance threshold judgment conditions, the vehicle's driving area can be accurately identified.

[0020] Further, determining the vehicle's current driving area also includes: if the link where the current position is located belongs to a first area type, and the vehicle's signal strength exceeds a preset signal strength threshold and the duration exceeds a first time threshold, then the vehicle's current driving area is determined to be the first area; if the link where the current position is located belongs to a first area type, and the link where the first position is located belongs to a second area type, and the height difference between the current position and the first position is less than a fourth preset height threshold, then the vehicle's current driving area is determined to be the first area.

[0021] In this solution, the current driving area of ​​the vehicle is determined as the first area by multi-dimensional data fusion (link type, signal strength and duration, height difference), which can significantly improve recognition accuracy, environmental adaptability and reliability, and at the same time provide key support for autonomous driving safety decisions.

[0022] Based on the same concept, a vehicle driving area identification system is provided. The system includes: a navigation and positioning module for locating the vehicle's position on an SD map; and a processing module for establishing a second intersection node queue based on the SD map, wherein the SD map includes a first region, a third region, and a second region connecting the first and third regions, and the second intersection node is an intersection node connecting the second region to the first region; when the vehicle is located to be driving to any node in the second node queue, the vehicle's position is recorded and designated as the first position; after the vehicle passes the first position, the current driving area of ​​the vehicle is determined based on the current position, the first position, and the vehicle positioning information; and the current driving area information is output.

[0023] Compared with the prior art, the beneficial effects of this application are as follows: This solution utilizes SD maps and spatial topology logic to replace existing methods that rely on pixel-level image data analysis for vehicle driving area identification, addressing the high resource consumption and low efficiency of existing solutions in specific scenarios. Specifically, this solution abandons traditional image recognition techniques and deep learning networks for real-time analysis of high-resolution images. Instead, it uses predefined SD maps and intersection node topology data for localization. The amount of structured information such as node details and connectivity is only a fraction of that of image data, reducing the parallel computing burden on GPUs / TPUs. Through relevant node queues, vehicles quickly record their positions when a node is triggered, avoiding latency issues in image recognition and improving system real-time performance. By leveraging the structured data of SD maps, rather than relying on image features, the robustness of vehicle driving area identification is improved, and the dependence on massive amounts of labeled data is reduced. Furthermore, rapid localization and position recording enable real-time determination of the driving area, providing timely input for vehicle control (such as path planning or speed adjustment), thereby enhancing safety performance. Therefore, the vehicle driving area recognition method based on SD map in this technical solution improves the recognition accuracy of vehicles in special areas (such as underground parking lots) and provides a feasible path for vehicle driving area recognition technology in intelligent driving technology. Attached Figure Description

[0024] Figure 1 This is an example diagram of the SD map described in this application.

[0025] Figure 2 This is a flowchart of the vehicle driving area identification method described in this application.

[0026] Figure 3 This is an example diagram of the first, second, and third regions established based on the SD map as described in this application.

[0027] Figure 4 This is a flowchart illustrating the process of establishing a second intersection node queue based on an SD map as described in this application.

[0028] Figure 5 This is a flowchart of the second node queue for searching the second region based on directed graph traversal as described in this application.

[0029] Figure 6 This is a flowchart illustrating the process of determining the second intersection node queue from the second node queue as described in this application.

[0030] Figure 7 This is a schematic diagram of the vehicle driving area identification system described in this application. Detailed Implementation

[0031] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0032] To better understand the technical content of the embodiments of this application, the technical terms involved in the embodiments of this application will be explained.

[0033] SD maps (Standard Definition Maps) are navigation maps with meter-level accuracy, such as... Figure 1 The image shown is an example of an SD map, primarily designed to provide basic route planning services for human drivers. It includes macro-level geographic information such as the road network framework, points of interest, and administrative boundaries, but lacks lane-level details (such as lane line type and curvature). Its production and maintenance costs are significantly lower than high-precision maps, and it updates more frequently. SD maps only display ground-level road network information and lack elevation information, making it impossible to distinguish between ground and underground information, road speed limits, etc.

[0034] In SD maps, a link refers to a complete road segment between two road intersections. For example, a straight road from intersection A to intersection B is a link. Link nodes mark points where roads intersect or their attributes change. The first and last nodes of a link are the start and end nodes, respectively. Links connect with other links through nodes, forming a complete road network topology. In SD maps, link types include provincial highways, national highways, expressways, urban roads, and regional connecting roads. Each link has its own attributes, including linkid, type, and length. Example 1

[0035] This embodiment provides a vehicle driving area recognition method, applicable to L2 / L3 level autonomous vehicles, unmanned delivery vehicles, unmanned trucks, and robots permitted to operate on roads. It achieves high-precision, low-computing-power vehicle driving area recognition. The driving area referred to in this method is the area connected to underground road space, which can be an underground parking lot, an underground tunnel, etc., without specific limitations. This method can accurately identify scenarios of entering and exiting underground road space. To more clearly illustrate the method described in this embodiment, the scenario of a vehicle entering and exiting an underground parking lot is used as an example. Currently, commonly used underground parking lot scene recognition methods rely on image recognition and deep learning networks. These methods rely on massive amounts of labeled images to train deep learning models (such as CNN, YOLO, etc.). Typically, a large amount of image data from the vehicle's camera is collected and input into a deep learning network to obtain an underground parking lot recognition network model. However, this method requires a large amount of data for training and cannot recognize certain underground parking lot scenarios that have not been trained, and its accuracy in underground parking lot scene recognition is not high enough.

[0036] This method, based on SD maps and vehicle location information, is an underground parking lot scene recognition approach that can accurately identify scenes of entering and exiting underground parking lots. For example... Figure 2 As shown, the specific method mainly includes steps S1 to S4.

[0037] S1. Establish a second intersection node queue based on the SD map. The SD map includes a first region, a third region, and a second region connecting the first and third regions. The second intersection node is the intersection connecting the second region to the first region.

[0038] S2, when the vehicle is located to be traveling to any node in the second intersection node queue, record the vehicle's position and mark it as the first position.

[0039] S3, after the vehicle passes the first location, determines the vehicle's current driving area based on the current location, the first location, and the vehicle's positioning information.

[0040] S4 outputs the current driving area information.

[0041] It should be noted that, as Figure 3As shown, the third area is an underground area, specifically an underground parking lot in this embodiment. The underground parking lot may have one or more entrances / exits. The SD map data only includes entrance / exit information for the third area. The first area is the main road area near the underground parking lot. The second area is the underground parking lot connection area, connecting the main road area and the underground parking lot. The underground parking lot connection area (second area) has multiple connecting roads, connecting the entrances / exits of the underground parking lot to the main road in the first area. Both the main roads and connecting roads in the first and second areas are linked. The second intersection node is the connection point between the main road area and the underground parking lot connection area.

[0042] like Figure 4 As shown, in step S1, a second intersection node queue is established based on the SD map, including: S11, detect all links within a preset range on the SD map centered on the vehicle, as well as the start and end nodes of the links; S12, construct a directed graph based on all links and the start and end nodes of each link; S13, based on the directed graph traversal, search for nodes located in the second region to form a second node queue, and determine the second intersection node queue from the second node queue.

[0043] It should be noted that in this embodiment, the SD map can be stored in the vehicle's local memory or in the cloud, and supports automatic or manual updates. It provides static road network data, and can be read directly from the local storage or the cloud when needed. Vehicle positioning information is determined through a combined navigation approach. Specifically, two or more heterogeneous navigation technologies (such as satellite navigation, inertial navigation, and radio navigation) are fused using a computer and data processing algorithm (the core of which is Kalman filtering) to form a more accurate and reliable navigation system. During vehicle operation, navigation data is received to determine vehicle positioning information, such as latitude, longitude, altitude, attitude angle, and the presence or absence of RTK (Real-Time Kinematic) signals. Using the vehicle positioning information and the SD map, the vehicle's location on the specific link in the SD map can be determined in real time using a Hidden Markov Chain algorithm.

[0044] Furthermore, the preset range can be dynamically determined according to the actual scenario. In this embodiment, the preset range is a radius of 1 kilometer centered on the vehicle. Within this range, all links in the first and second regions involved, as well as the start and end nodes of the links, can be detected. When establishing a directed graph, such as... Figure 3As shown, an SD map with a radius of 1km centered on the vehicle is read. The start and end nodes are treated as nodes in a directed graph, links are treated as edges, and the length of a link is treated as the length of an edge. The directed graph has direction; travel is only possible from the start node to the end node, and a return trip is not possible. A second intersection node is searched by traversing the directed graph. This second intersection node serves as a key geofence node connecting different areas, providing a static reference for the vehicle's location. Establishing a queue of second intersection nodes is the core foundation for improving the accuracy and efficiency of area recognition.

[0045] like Figure 5 As shown, in step S13, the nodes located in the second region are searched based on the directed graph traversal to form a second node queue, including: S131, Establish and initialize a search queue based on the starting node of the link where the vehicle is located, where the starting node of the link where the vehicle is located is the head node of the search queue. S132, sequentially extract the search nodes in the search queue, and determine the first end node of all outgoing edges of each search node; S133, If the first end node belongs to the second region node and the first end node has not been extracted, add the first end node to the search queue as a search node, and add the first end node to the second node queue as a second node. S134: Completely traverse the search nodes and output the second node queue.

[0046] It's important to note that directed graph traversal search is a core method in vehicle driving area identification, used to model road networks. Nodes represent intersections or locations, and directed edges represent driving directions. Road networks inherently possess graph structure characteristics, and using a directed graph (with edge directions) can accurately represent road connectivity and driving restrictions. During the traversal process, after identifying the first region, a dedicated node queue (i.e., the second node queue) is maintained to process nodes in the second region. This can be implemented using a breadth-first search (BFS) algorithm. By separating queue management, region identification efficiency is optimized, significantly improving computational efficiency and real-time performance.

[0047] After the loop traversal search is completed, all nodes in the second region can be found, forming the second node queue. For example... Figure 6 As shown, determining the second intersection node queue from the second node queue includes: S135, sequentially extract the second nodes from the second node queue, and determine the second end node of all outgoing edges of each second node; S136, If the second end node belongs to the second node queue, the second end node has a successor link, and the second end node has not been extracted, add the second end node to the second intersection node queue as the second intersection node. S137, until the second end node is completely traversed, output the second intersection node queue.

[0048] Based on a directed graph, a breadth-first search is used for traversal. Only intersection nodes connecting the second region to the first region are selected, while irrelevant road nodes (such as internal passages) are excluded. This reduces the computational load for subsequent region determinations, accurately locates key nodes, and avoids interference from redundant data. After the iterative traversal search is completed, all second intersection nodes are found, forming a queue of second intersection nodes.

[0049] In step S2, when the vehicle is located to be traveling to any node in the second intersection node queue, the vehicle's position is recorded and designated as the first position. Specifically, the first position is the intersection leading into the second area. The location information of the first position, including longitude, latitude, and altitude, can be determined through integrated navigation.

[0050] In step S3, after the vehicle passes the first location, the current driving area of ​​the vehicle is determined based on the current location, the first location, and the vehicle positioning information, including: The vehicle's operating area includes Zone 1, Zone 2, and Zone 3; If the height difference between the current position and the first position is greater than the first preset height threshold, the distance between the current position and the first position is less than the first preset distance threshold, and the vehicle signal strength does not exceed the preset signal strength threshold, then it is determined that the vehicle has entered the third area. If the distance between the current position and the first position exceeds the first preset distance threshold, and the link where the current position is located belongs to the second region type, the current position is re-recorded as the first position.

[0051] After the vehicle enters the second area, to accurately determine its current position during continued travel, a first position is used as a reference. The distance difference, altitude difference, and positioning signal characteristics between the current and first positions are used to comprehensively determine whether the vehicle has entered a specific third area. This multi-dimensional data fusion judgment improves the accuracy and reliability of boundary determination compared to relying solely on traditional planar positioning. Altitude difference information provides a crucial third-dimensional criterion, effectively resolving ambiguities in the vertical direction. Distance difference provides a more accurate relative positional relationship than latitude and longitude alone. Positioning signal characteristics provide auxiliary evidence of entering a specific environmental type.

[0052] It should be noted that the vehicle positioning information in this embodiment can use RTK signals. RTK signals are high-precision positioning enhancement signals based on carrier phase differential technology. Through the coordinated work of the base station and the rover, satellite positioning errors are eliminated, achieving real-time positioning. There are no RTK signals in areas with obstructed satellite signals, and there are no RTK signals when the vehicle is parked indoors or underground. In other embodiments, other positioning signals can also be used for vehicle positioning information, such as LBS signals (Location-Based Service, also known as base station positioning), and no specific limitation is made here.

[0053] Taking entering an underground parking lot as an example, the first preset height threshold and the first preset distance threshold are set according to the specific actual situation and are not specifically limited here. In this embodiment, the first preset height threshold is 3m, the first preset distance threshold is 150m, and the preset signal strength threshold is 0 (indicating no RTK signal). When a vehicle enters the underground parking lot connection area from the second intersection node, the vehicle's current position is located in real time, the link the vehicle is currently in is determined, and the type of link is determined. During the vehicle's journey, if the height difference between the current position and the first position is greater than 3m, the distance between them is less than 150m, and the vehicle continuously has no RTK signal, it indicates that the vehicle has entered the underground parking lot. In addition, during the vehicle's journey, when the distance between the current position and the second intersection node reaches 150m, this position is re-recorded and defined as the first position, thereby ensuring the accuracy of subsequent judgments. It should be noted that the height difference and distance between the current position and the first position can be calculated using the latitude and longitude of the current position and the latitude and longitude of the first position.

[0054] Furthermore, in step S3, determining the vehicle's current driving area also includes: If the distance between the current location and the first location exceeds the first preset distance threshold, the link where the current location is located belongs to the first region type, or the vehicle signal strength exceeds the preset signal strength threshold and the duration exceeds the first time threshold, then the vehicle's current driving area is confirmed as the first region.

[0055] Taking the example of not entering the underground parking lot, the first time threshold is set according to the specific actual situation and is not specifically limited here. In this embodiment, the first time threshold is 6 seconds. After the vehicle passes the second intersection node, it does not enter the second area. By fusing multi-dimensional data such as distance difference, link type, signal strength, and duration, the accuracy and scene adaptability of identifying the vehicle entering the first area after passing the second intersection node can be significantly improved. This is especially suitable for autonomous driving or intelligent navigation scenarios in complex road environments.

[0056] Furthermore, in step S3, determining the vehicle's current driving area also includes: If the vehicle enters the third area, determine the link ID (linkid) of the last link the vehicle entered the third area. rec And re-record the position of entering the third area as the first position; The third area includes multiple entrances and exits. When a vehicle leaves the third area, the link ID (linkid) of the current location is used to pinpoint the location. cur ; If linkid cur ≠linkid rec If the height difference between the current position and the first position is less than the second preset height threshold, and the link where the current position is located belongs to the second region type, then the current driving region of the vehicle is determined to be the second region, and the slot_flag flag of the vehicle's third region detection is set to true. If linkid cur =linkid rec If the height difference between the current position and the first position is less than the third preset height threshold, and the link where the current position is located belongs to the second region type, then the current driving area of ​​the vehicle is determined to be the second region, and the slot_flag flag of the vehicle's third region detection is set to true.

[0057] Taking the example of a vehicle entering and exiting the underground parking lot via a connecting area, the second and third preset height thresholds are set according to specific circumstances and are not specifically limited here. In this embodiment, the second preset height threshold is 3m and the third preset height threshold is 2m. The linkid of the vehicle entering the underground parking lot is determined... rec With Linkid exiting the underground parking lot cur Whether they are the same determines whether they exited from the same or different entrances / exits. cur ≠linkid rec Indicates exiting from different entrances / exits, linkid cur =linkid rec This indicates that the vehicle exited from the same entrance / exit. After determining whether the vehicle exited from the same or different entrance / exit, the vehicle's driving area can be determined by the height difference between the current location and the first location, and the type of the link to which the current location belongs. This improves the accuracy and scene adaptability of the second area recognition.

[0058] Furthermore, in step S3, determining the vehicle's current driving area also includes: If the vehicle's third region detection success flag slot_flag is false, the link where the current position is located belongs to the first region type, the link where the first position is located belongs to the second region type, and the distance between the current position and the first position exceeds the second preset distance threshold, then the vehicle's current driving region is determined to be the first region. If the vehicle's third region detection success flag (slot_flag) is true, the link where the current position is located belongs to the first region type, and the link where the first position is located belongs to the second region type, then the vehicle's current driving region is determined to be the first region.

[0059] The vehicle's third-region detection success flag, `slot_flag`, defaults to false. After a vehicle passes the second intersection node and enters the second region, if `slot_flag` is false, it indicates the vehicle has not entered the third region; if `slot_flag` is true, it indicates the vehicle has entered and exited the third region. By determining whether the vehicle has entered the third region, and combining the link type and distance threshold conditions of the first and current positions, the vehicle's travel area can be accurately identified.

[0060] Specifically, taking entering and exiting an underground parking lot as an example, if the vehicle's underground parking lot detection success flag (slot_flag) is false, it indicates that the vehicle has not entered the underground parking lot. The link at the current location belongs to the first area type, the link at the first location belongs to the second area type, and the distance between the current location and the first location exceeds a second preset distance threshold. This means that after entering the underground parking lot connecting area, the vehicle did not enter the underground parking lot but instead left the underground parking lot connecting area and entered the first area. In this embodiment, the second preset distance threshold is 30m. In other possible embodiments, the second preset distance threshold can be other values, which are not specifically limited here. If the vehicle's underground parking lot detection success flag (slot_flag) is true, it indicates that the vehicle has entered and exited the underground parking lot. At this time, the vehicle will definitely enter the first area. Therefore, by determining that the link at the current location belongs to the first area type and the link at the first location belongs to the second area type, it can be determined that the vehicle is currently in the first area.

[0061] Furthermore, in step S3, determining the vehicle's current driving area also includes: After the vehicle exited the third area, If the link where the current location is located belongs to the first region type, and the vehicle signal strength exceeds the preset signal strength threshold and the duration exceeds the first time threshold, then the vehicle's current driving area is determined to be the first region. If the link where the current location is located belongs to the first region type, and the link where the first location is located belongs to the second region type, and the height difference between the current location and the first location is less than the fourth preset height threshold, then the current driving area of ​​the vehicle is determined to be the first region.

[0062] When the vehicle is on the ground, it has an RTK signal. Therefore, after the vehicle leaves the third area, if the link where the vehicle was located before positioning belongs to the first area type, and the vehicle's signal strength exceeds a preset signal strength threshold and the duration exceeds a first time threshold, the vehicle's current driving area can be accurately determined to be the first area. Additionally, in this embodiment, the fourth preset height threshold is 2m. In other possible embodiments, the second preset distance threshold can be other values, which are not specifically limited here. After the vehicle leaves the third area, by judging the height difference between the vehicle's current position and the first position, as well as the link type, the vehicle's current driving area can also be accurately determined to be the first area.

[0063] It should be noted that, in order to improve the efficiency of vehicle driving area recognition, a parallel judgment method is adopted in the above judgment conditions. Among the different judgment conditions for vehicles driving in the same area, it is sufficient to meet only one of them.

[0064] In this embodiment, a vehicle driving area identification method based on SD maps replaces the existing technology's pixel-level analysis of image data with spatial topology logic, addressing the high cost and low efficiency of existing solutions in identifying special scenarios. Specifically, this solution abandons traditional image recognition techniques and deep learning networks for real-time analysis of high-resolution images, instead using predefined SD maps and intersection node topology data for localization. The amount of structured information such as node information and connection relationships is only one-thousandth of that of image data, reducing the parallel computing pressure on GPUs / TPUs. Through relevant node queues, vehicles quickly record their positions when a node is triggered, avoiding the latency problem in image recognition and improving system real-time performance. By utilizing the structured data of SD maps, rather than relying on image features, the robustness of vehicle driving area identification is improved, and the dependence on massive amounts of labeled data is reduced. Furthermore, through rapid localization and position recording, real-time determination of the driving area is achieved, providing timely input for vehicle control (such as path planning or speed adjustment), thereby enhancing safety performance. Therefore, the vehicle driving area recognition method based on SD map in this technical solution improves the recognition accuracy of vehicles in special areas (such as underground parking lots) and provides a feasible path for vehicle driving area recognition technology in intelligent driving technology. Example 2

[0065] like Figure 7 As shown, this application also provides a vehicle driving area identification system, which includes a navigation and positioning module 100 and a processing module 200.

[0066] The navigation and positioning module 100 is used to locate the vehicle's position on an SD map. It may include a satellite signal receiving antenna, a satellite signal receiving circuit, an inertial measurement unit (IMU), sensors such as a vehicle speed pulse sensor, communication circuitry (e.g., cellular network circuitry), a memory storing SD map data and a program for vehicle positioning and navigation based on the output signals of the satellite signal receiving circuit and other sensor output signals, and a processor. The navigation and positioning module 100 may also include human-computer interaction modules such as a display, speaker, microphone, touchscreen, and knob, used to output vehicle positioning and navigation information to the user or receive user commands.

[0067] The processing module 200 establishes a second intersection node queue based on the SD map. The SD map includes a first region, a third region, and a second region connecting the first and third regions. The second intersection node is the intersection node connecting the second region to the first region. When a vehicle is located to be traveling to any node in the second node queue, the vehicle's position is recorded and designated as the first position. After the vehicle passes the first position, the current driving area of ​​the vehicle is determined based on the current position, the first position, and the vehicle's positioning information. The current driving area information is then output.

[0068] Although exemplary embodiments have been described herein with reference to the accompanying drawings, it should be understood that the above exemplary embodiments are merely illustrative and are not intended to limit the scope of this application. Various changes and modifications can be made therein by those skilled in the art without departing from the scope and spirit of this application. All such changes and modifications are intended to be included within the scope of this application as claimed in the appended claims.

[0069] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0070] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed.

[0071] The various component embodiments of this application can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) can be used in practice to implement some or all of the functions of some modules according to the embodiments of this application. This application can also be implemented as an apparatus program (e.g., a computer program and computer program product) for performing part or all of the methods described herein. Such an implementation of this application can be stored on a computer-readable medium, or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0072] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0073] Although the description of this application has been made in conjunction with the specific embodiments described above, it will be apparent to those skilled in the art that many substitutions, modifications, and variations can be made based on the foregoing. Therefore, all such substitutions, modifications, and variations are included within the spirit and scope of the appended claims.

Claims

1. A method for identifying vehicle driving areas, characterized in that, The method includes: A second intersection node queue is established based on the SD map, which includes a first region, a third region, and a second region connecting the first region and the third region. The second intersection node is the intersection connecting the second region to the first region. When a vehicle is located to be traveling to any node in the second intersection node queue, the vehicle's position is recorded and designated as the first position; After the vehicle passes the first location, the current driving area of ​​the vehicle is determined based on the current location, the first location, and the vehicle positioning information. Output the current driving area information.

2. The vehicle driving area identification method according to claim 1, characterized in that, The process of establishing a second intersection node queue based on the SD map includes: The detection is centered on the vehicle and includes all links within a preset range on the SD map, as well as the start and end nodes of each link. Construct a directed graph based on all the links, as well as the start and end nodes of each link; Based on the directed graph traversal search, a second node located in the second region is formed, a second node queue is formed, and the second intersection node queue is determined from the second node queue.

3. The vehicle driving area identification method according to claim 2, characterized in that, The step of searching for a second node located in the second region based on the directed graph traversal to form a second node queue includes: A search queue is established and initialized based on the starting node of the link where the vehicle is located, and the starting node of the link where the vehicle is located is the head node of the search queue. Extract the search nodes from the search queue in sequence, and determine the first end node of all outgoing edges of each search node; If the first ending node belongs to the second region node and the first ending node has not been extracted, add the first ending node to the search queue as the search node, and add the first ending node to the second node queue as the second node. Completely traverse the search nodes and output the second node queue.

4. The vehicle driving area identification method according to claim 3, characterized in that, Determining the second intersection node queue from the second node queue includes: Extract the second nodes from the second node queue in sequence, and determine the second end node of all outgoing edges of each second node; If the second end node belongs to the second node queue, the second end node has a successor link, and the second end node has not been extracted, add the second end node to the second intersection node queue as the second intersection node; Completely traverse the second ending node and output the queue of the second intersection nodes.

5. The vehicle driving area identification method according to claim 1, characterized in that, After the vehicle passes the first location, determining the vehicle's current driving area based on the current location, the first location, and the vehicle's positioning information includes: If the height difference between the current position and the first position is greater than a first preset height threshold, the distance between the current position and the first position is less than a first preset distance threshold, and the vehicle signal strength does not exceed a preset signal strength threshold, then the current driving area of ​​the vehicle is determined to be the third area. If the distance between the current position and the first position exceeds the first preset distance threshold, and the link where the current position is located belongs to the second region type, the current position is re-recorded as the first position.

6. The vehicle driving area identification method according to claim 5, characterized in that, Determining the vehicle's current driving area also includes: If the distance between the current location and the first location exceeds a first preset distance threshold, the link where the current location is located belongs to a first region type, and the vehicle signal strength exceeds a preset signal strength threshold and the duration exceeds a first time threshold, then the current driving area of ​​the vehicle is determined to be the first region.

7. The vehicle driving area identification method according to claim 6, characterized in that, Determining the vehicle's current driving area also includes: If the vehicle enters the third area, determine the link ID (linkid) of the last link the vehicle entered the third area. rec And re-record the position of entering the third region as the first position; The third area includes multiple entrances and exits. When a vehicle leaves the third area, the link number (linkid) of the current location is determined. cur ; If linkid cur ≠linkid rec If the height difference between the current position and the first position is less than the second preset height threshold, and the link where the current position is located belongs to the second region type, then the current driving region of the vehicle is determined to be the second region, and the third region detection success flag of the vehicle is set to true. If linkid cur =linkid rec If the height difference between the current position and the first position is less than the third preset height threshold, and the link where the current position is located belongs to the second region type, then the current driving area of ​​the vehicle is determined to be the second region, and the third region detection success flag of the vehicle is set to true.

8. The vehicle driving area identification method according to claim 7, characterized in that, Determining the vehicle's current driving area also includes: If the third region detection success flag of the vehicle is false, the link where the current position is located belongs to the first region type, the link where the first position is located belongs to the second region type, and the distance between the current position and the first position exceeds the second preset distance threshold, then the current driving area of ​​the vehicle is determined to be the first region. If the third region detection success flag of the vehicle is true, the link where the current position is located belongs to the first region type, and the link where the first position is located belongs to the second region type, then the current driving region of the vehicle is determined to be the first region.

9. The vehicle driving area identification method according to claim 7, characterized in that, Determining the vehicle's current driving area also includes: If the link where the current location is located belongs to the first region type, and the vehicle signal strength exceeds the preset signal strength threshold and the duration exceeds the first time threshold, then the current driving area of ​​the vehicle is determined to be the first region. If the link where the current position is located belongs to the first region type, and the link where the first position is located belongs to the second region type, and the height difference between the current position and the first position is less than the fourth preset height threshold, then the current driving area of ​​the vehicle is determined to be the first region.

10. A vehicle driving area identification system, characterized in that, The system includes: The navigation and positioning module is used to locate the vehicle's position on the SD map; The processing module is used to establish a second intersection node queue based on the SD map. The SD map includes a first region, a third region, and a second region connecting the first region and the third region. The second intersection node is the intersection node connecting the second region to the first region. When a vehicle is located to be traveling to any node in the second node queue, the vehicle's position is recorded and designated as the first position. After the vehicle passes the first position, the current driving area of ​​the vehicle is determined based on the current position, the first position, and the vehicle's positioning information. The current driving area information is then output.

Images