Lane determination method and device and electronic device

Abstract

The invention discloses a lane determination method and device and an electronic device. The method comprises the following steps: acquiring target standard precise map data and target sensing data; an initial local map is generated based on the target sensing data, and the initial local map is used for describing the road topological relation of the road where the target vehicle is located; the initial local map is corrected based on the historical perception data, the target perception data and the target standard precise map data, and a target local map is obtained and comprises at least one drivable lane; a target lane is determined from the at least one drivable lane based on a preset road scoring strategy, the target lane is displayed in the display interface, and the preset road scoring strategy is used for scoring any drivable road according to at least one scoring rule. According to the invention, the technical problem of how to provide accurate and real-time lane recommendation service under the condition of no high-precision map data so as to improve the driving efficiency, reduce the cost and improve the adaptability is solved.

Description

Technical Field

[0001] The present invention relates to the field of vehicle driving technology, and in particular to a lane determination method, device and electronic device. Background Art

[0002] With the development of autonomous driving technology, the safety and comfort of autonomous vehicles are becoming increasingly important. Lane recommendation provides lane references for autonomous vehicles, preventing them from straying from their lanes and ensuring safe driving.

[0003] At present, lane recommendation systems are mainly based on high-precision map data. By analyzing high-precision map data, lanes and road features are identified to provide lane references for autonomous vehicles. In autonomous driving systems, high-precision map data can provide lane recommendations for autonomous vehicles in advance, but high-precision map data requires sophisticated equipment and high-cost production methods, and the update cycle of high-precision map data is long. In order to ensure the safety of autonomous vehicles, autonomous vehicles need to collect road and obstacle information during driving and build local maps, but they cannot capture real-time changes in the road in a timely manner, which may cause deviations in recommended lanes and affect the safety of autonomous vehicles. Therefore, how to improve the accuracy of lane recommendations for autonomous vehicles and ensure the safety of autonomous vehicles is a problem that those skilled in the art currently need to solve.

[0004] To address the above-mentioned problems, no effective solutions have been proposed so far. Summary of the Invention

[0005] Embodiments of the present invention provide a lane determination method, device, and electronic device to at least solve the technical problem of how to provide accurate and real-time lane recommendation services in the absence of high-precision map data to improve driving efficiency while reducing costs and improving adaptability.

[0006] According to one aspect of an embodiment of the present invention, a lane determination method is provided, comprising: acquiring target precise map data and target perception data, wherein the target perception data is used to represent road information of a road where a target vehicle is located; generating an initial local map based on the target perception data, wherein the initial local map is used to describe the road topology relationship of the road where the target vehicle is located; correcting the initial local map based on historical perception data, the target perception data, and the target precise map data to obtain a target local map, wherein the target local map includes at least one drivable lane; determining a target lane from the at least one drivable lane based on a preset road scoring strategy to display the target lane in a display interface, wherein the preset road scoring strategy is used to score any drivable road according to at least one scoring rule.

[0007] Furthermore, the target local map also includes at least one of the following: lane number information, lane type information, lane drivable information, and obstacle ratio information on the lane.

[0008] Furthermore, obtaining target perception data includes: obtaining initial perception data; performing data preprocessing on the initial perception data to obtain preprocessed perception data; and performing data labeling on the preprocessed perception data to obtain target perception data.

[0009] Furthermore, generating an initial local map based on the target perception data includes: determining the location information of the target vehicle; dividing multiple roads based on intersection data in the target perception data; dividing multiple lanes based on lane line data in the target perception data; and generating an initial local map based on the location information, multiple roads, and multiple lanes.

[0010] Furthermore, obtaining target precise map data includes: obtaining initial precise map data, wherein the initial precise map data is map data in a world coordinate system; converting the initial precise map data into a vehicle coordinate system to obtain target precise map data.

[0011] Furthermore, the initial local map is corrected based on the historical perception data, the target perception data and the target precise map data to obtain the target local map, including: matching the target perception data and the target precise map data to obtain an initial matching result; in response to the matching result including matching failure information, determining the preceding perception data of the preceding image frame based on the historical perception data, wherein the timestamp corresponding to the preceding image frame is earlier than the timestamp corresponding to the current image frame, and the target perception data is the perception data of the current image frame; correcting the matching failure information based on the confidence of the preceding perception data and the confidence of the target precise map data to obtain matching success information; updating the initial matching result based on the matching success information to obtain the target matching result; and correcting the initial local map based on the target matching result to obtain the target local map.

[0012] Furthermore, determining a target lane from at least one drivable lane based on a preset road scoring strategy includes: determining navigation recommended lane information based on target landmark map data; scoring at least one drivable lane based on the preset road scoring strategy, navigation recommended lane information and target perception data to obtain a lane score value; and determining a target lane from at least one drivable lane based on the lane score value.

[0013] Furthermore, at least one drivable lane is scored based on a preset road scoring strategy, navigation recommended lane information and target perception data, and the lane score value obtained includes: obtaining the distance between any drivable lane and the target vehicle, the volume ratio of obstacles on any drivable lane, and the historical driving data of any drivable lane based on the target perception data; determining a first score of any drivable lane based on the distance, determining a second score of any drivable lane based on the volume ratio, determining a third score of any drivable lane based on the navigation recommended lane information, and determining a fourth score of any drivable lane based on the historical driving data; determining a lane score value of any drivable lane based on the first score, the second score, the third score and the fourth score.

[0014] According to another aspect of an embodiment of the present invention, a lane determination device is also provided, including: an acquisition module for acquiring target landmark map data and target perception data, wherein the target perception data is used to represent road information of the road where the target vehicle is located; a generation module for generating an initial local map based on the target perception data, wherein the initial local map is used to describe the road topology relationship of the road where the target vehicle is located; a correction module for correcting the initial local map based on historical perception data, target perception data and target landmark map data to obtain a target local map, wherein the target local map includes at least one drivable lane; a determination module for determining a target lane from at least one drivable lane based on a preset road scoring strategy to display the target lane in a display interface, wherein the preset road scoring strategy is used to score any drivable road according to at least one scoring rule.

[0015] Optionally, the target local map further includes at least one of the following: lane number information, lane type information, lane drivable information, and obstacle ratio information on the lane.

[0016] Optionally, the acquisition module is further used to acquire initial perception data; perform data preprocessing on the initial perception data to obtain preprocessed perception data; and perform data labeling on the preprocessed perception data to obtain target perception data.

[0017] Optionally, the generation module is also used to determine the location information of the target vehicle; divide multiple roads based on intersection data in the target perception data; divide multiple lanes based on lane line data in the target perception data; and generate an initial local map based on the location information, multiple roads and multiple lanes.

[0018] Optionally, the acquisition module is also used to obtain initial precise map data, wherein the initial precise map data is map data in a world coordinate system; the initial precise map data is converted to the vehicle coordinate system to obtain target precise map data.

[0019] Optionally, the correction module is also used to match the target perception data and the target precision map data to obtain an initial matching result; in response to the matching result including matching failure information, determine the previous perception data of the previous image frame based on the historical perception data, wherein the timestamp corresponding to the previous image frame is earlier than the timestamp corresponding to the current image frame, and the target perception data is the perception data of the current image frame; based on the confidence of the previous perception data and the confidence of the target precision map data, the matching failure information is corrected to obtain matching success information; based on the matching success information, the initial matching result is updated to obtain the target matching result; based on the target matching result, the initial local map is corrected to obtain the target local map.

[0020] Optionally, the determination module is also used to determine navigation recommended lane information based on target map data; score at least one drivable lane based on a preset road scoring strategy, navigation recommended lane information and target perception data to obtain a lane scoring value; and determine the target lane from at least one drivable lane based on the lane scoring value.

[0021] Optionally, the determination module is also used to obtain the distance between any drivable lane and the target vehicle, the volume ratio of obstacles on any drivable lane, and the historical driving data of any drivable lane based on the target perception data; determine the first score of any drivable lane based on the distance, determine the second score of any drivable lane based on the volume ratio, determine the third score of any drivable lane based on the navigation recommended lane information, and determine the fourth score of any drivable lane based on the historical driving data; determine the lane score value of any drivable lane based on the first score, the second score, the third score and the fourth score.

[0022] According to another aspect of an embodiment of the present invention, a vehicle is further provided, comprising a computer program, which implements the methods in various embodiments of the present invention when executed by a processor.

[0023] According to another aspect of an embodiment of the present invention, an electronic device is provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods of various embodiments of the present invention when running.

[0024] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the methods in various embodiments of the present invention.

[0025] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the method in each embodiment of the present invention is implemented.

[0026] In this embodiment of the present invention, a method is employed to integrate high-definition map data with autonomous driving perception data. First, on-board sensors collect real-time environmental information about the target vehicle's roadway to construct an initial local map. This map depicts the current road topology, including lane layout and connectivity. This initial map is then corrected using historical perception data and target high-definition map data to form a target local map. This ensures that it includes detailed information about at least one drivable lane and corrects errors caused by temporary changes. Finally, based on a pre-defined road scoring strategy, all drivable lanes are scored based on multiple factors, including lane conditions, obstacle distribution, navigation guidance, and user preferences. The highest-scoring lane is selected as the target lane and intuitively displayed on the vehicle's display interface, guiding the driver to make the optimal lane selection. This reduces the user's reliance on high-definition map data, thereby achieving a low-cost, highly adaptable, and real-time lane recommendation service. This addresses the technical challenge of providing accurate, real-time lane recommendation services in the absence of high-precision map data, thereby improving driving efficiency while reducing costs and increasing adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0028] Figure 1 is a flow chart of a lane determination method according to an embodiment of the present invention;

[0029] Figure 2 is a flowchart of a lane recommendation method according to an embodiment of the present invention;

[0030] Figure 3 is a flowchart of a scoring strategy according to an embodiment of the present invention;

[0031] Figure 4 4 is a module diagram of a lane determination device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0033] In order to enable those skilled in the art to better understand the present invention, some glossaries related to the present invention are given below.

[0034] Standardized map data: Standardized map data is a level of accuracy between standard and high-precision maps. It contains basic road information, such as road name, type, and number of lanes, but lacks the location accuracy and detail of high-precision maps. While suitable for general navigation services, it may not be accurate enough for applications such as autonomous driving that require high-precision positioning.

[0035] Autonomous driving perception data: Autonomous driving perception data is real-time data about the surrounding environment collected by various sensors installed on the vehicle (such as cameras, lidar, and millimeter-wave radar). This data includes information such as obstacles, traffic signs, and road conditions around the vehicle, and is a key basis for autonomous vehicles to make driving decisions.

[0036] Lane Recommendation Service: This service recommends the most suitable lane for drivers based on their current vehicle location, navigation destination, and real-time traffic conditions. This service helps drivers avoid entering the wrong lane, improving driving efficiency and road use, and is particularly effective in complex road environments.

[0037] Lateral positioning: Lateral positioning refers to the technology that determines the lateral position of a vehicle on the road, that is, its position relative to the lane markings. This is crucial for autonomous vehicles, ensuring that the vehicle stays in the correct lane and avoids deviating from the lane markings.

[0038] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0039] According to an embodiment of the present invention, an embodiment of a lane determination method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0040] An embodiment of the present application provides a lane determination method. The lane determination method can be used to provide a lane recommendation function for a preset application scenario. The above-mentioned preset application scenarios may include the following scenarios in the vehicle field: commuting autonomous driving scenarios, artificial intelligence (AI) driving scenarios for family cars, automatic parking assistance (APA) scenarios (such as memory parking for owned parking spaces in garages, smart parking for designated parking spaces in parking lots, etc.), and intelligent navigation assistance (Navigation Guided Pilot, NGP) scenarios in urban areas or highway areas.

[0041] Figure 1 FIG. 1 is a flow chart of a lane determination method according to an embodiment of the present invention. Figure 1 As shown, the lane determination method includes the following steps:

[0042] Step S10: Acquire target map data and target perception data, wherein the target perception data is used to represent road information of the road where the target vehicle is located.

[0043] In an embodiment of the present invention, the target precision map data can be understood as standard precision map data. Compared with high-precision maps, the precision of standard precision maps is lower, and usually contains basic road network information, such as the location of the road, the type of road (highway, urban road, etc.), the number of lanes, and the type of lane (such as straight lanes, turning lanes, etc.). Although the precision map data does not contain detailed information at the lane level, such as the precise location of lane lines, its data is relatively easy to update, has a wide coverage, and can provide the basic topological structure of the road. For example, in the present invention, the target precision map data is used to provide a basic information framework for the vehicle's driving road, and is used in combination with real-time perception data to achieve lane recommendations, which is not limited here.

[0044] Target perception data can be understood as the vehicle's surrounding environment information collected and processed in real time by autonomous vehicles through on-board sensors (such as cameras, lidars, millimeter-wave radars, etc.). Exemplarily, target perception data includes but is not limited to road shape, lane lines, traffic signs, obstacles (such as pedestrians, vehicles, construction areas, traffic cones, etc.), and road conditions (such as wet, damaged, etc.). The perception data takes the target vehicle (i.e., the vehicle itself) as the origin, describes the position and state of the surrounding environment relative to the vehicle, and can reflect the real-time changes in the road and the current road conditions. In the present invention, target perception data is used to capture the real-time situation of the road, correct possible deviations in the precision map data, and ensure that lane recommendations are based on the current actual road conditions, which are not limited here.

[0045] Obtain target standard precision map data and target perception data. The target perception data is used to represent the road information of the target vehicle's road. It can be understood as obtaining standard precision map information corresponding to the current position of the target vehicle from the vehicle navigation system or map service provider, and obtaining detailed information collected in real time by various sensors on the target vehicle to capture the surrounding environment of the target vehicle, thereby providing real-time and specific reference information for lane recommendation.

[0046] In an embodiment of the present invention, target precise map data and target perception data are obtained. The target precise map data provides basic information and a macro-framework of the target vehicle's driving road, while the target perception data provides the real-time status and specific details of the target vehicle's surrounding environment. Lane recommendations based on real-time road conditions can be achieved, which can not only utilize the wide coverage of the precise map, but also ensure the accuracy and safety of the recommendations, and adapt to driving needs under various road conditions.

[0047] Step S11: generating an initial local map based on the target perception data, wherein the initial local map is used to describe the road topology relationship of the road where the target vehicle is located.

[0048] In this embodiment of the present invention, the initial local map can be understood as a local map of a section of road centered on the target vehicle (i.e., the ego vehicle), generated based on real-time environmental data collected by onboard sensors (such as cameras and lidar) during vehicle travel. For example, the initial local map includes the main components of the road, such as lane markings, roadside obstacles, traffic signs, and signal lights, and describes the position and status of these components relative to the vehicle, with the vehicle itself as the reference origin. This is not a limitation here.

[0049] Road topology can be understood as describing the spatial connectivity between roads and lanes. For example, in the context of autonomous driving and lane recommendation, road topology includes lane connectivity (e.g., straight ahead, left turn, right turn, lane change), lane type (e.g., straight ahead lane, turning lane, bus lane), and lane-to-intersection connectivity, among other things. These are not limited here.

[0050] Generating an initial local map based on target perception data can be understood as leveraging the environmental information captured in real time by sensors on the target vehicle (e.g., cameras, lidar, etc.) to construct a local map representation centered on the target vehicle that reflects the surrounding road structure and obstacle distribution. For example, the vehicle can identify and map the location and type of lane lines within a few meters ahead, static obstacles on the road (e.g., traffic signs, curbs, etc.), and dynamic obstacles (e.g., pedestrians, other vehicles, etc.), thereby better understanding its driving path and surrounding environment. This is not a limitation here.

[0051] The initial local map is used to describe the road topology of the road where the target vehicle is located. It can be understood that the initial local map not only contains the basic components of the road around the target vehicle, such as lanes, curbs, traffic signs, etc., but more importantly, the initial local map also reflects the logical connection and spatial layout between the above elements. For example, the road topology pays special attention to the relationship between lanes (such as straight, turning, bifurcation, merging, etc.), as well as the connection between lanes and road nodes such as intersections, exits, and entrances. For example, when a vehicle approaches a bifurcated intersection, the topological information in the local map can help the vehicle decide which lane is the best path to the destination and how to prepare for lane changes in advance, which is not limited here.

[0052] In this embodiment of the present invention, the initial local map generated based on target perception data is a dynamic, real-time, and vehicle-centric map model. This map details the current state of the vehicle's surroundings. The road topology describing the target vehicle's road builds on this model, further clarifying the logical relationships and spatial connections between various road sections, providing crucial foundational information for vehicle driving decisions.

[0053] Step S12: correcting the initial local map based on the historical perception data, the target perception data, and the target precise map data to obtain a target local map, wherein the target local map includes at least one drivable lane.

[0054] In the embodiments of the present invention, historical perception data can be understood as a collection of environmental information acquired and recorded by onboard sensors during the target vehicle's past travel. For example, historical perception data may include road characteristics, obstacle distribution, traffic flow conditions, and the like along the vehicle's route. Utilizing historical perception data can compensate for potential deficiencies in target perception data. For example, when the target vehicle is in a scenario with limited field of view (such as in a tunnel or behind a tall building), real-time perception information may not be comprehensive or accurate. In this case, historical perception data can serve as a supplement to help the vehicle better understand and predict current road conditions. This is not a limitation here.

[0055] The target local map can be understood as a revised local map representation that contains the precise topological relationship of the target vehicle's driving road and the information of the drivable lanes. For example, the target local map not only integrates the road conditions captured in real time in the target perception data, but also takes into account the long-term environmental characteristics recorded in the historical perception data, as well as the basic road information in the target precision map data. For example, in the target local map, each drivable lane is clearly marked, including the type of lane (such as straight, left turn, right turn), lane width, lane line shape, and information about possible obstacles on the lane, etc., which are not limited here.

[0056] A drivable lane can be understood as a lane in the target local map that is deemed unobstructed, free of obstacles, in compliance with traffic regulations, and safe for driving based on a comprehensive assessment of real-time perception information, historical data, and precise map information. For example, if the target perception data captures a temporary construction zone or heavy traffic congestion on a lane, then that lane will not be considered a drivable lane. Conversely, if a lane is free of obstacles, traffic flows smoothly, and the precise map data indicates that this lane is a recommended lane for the target vehicle's route, then that lane will be marked as a drivable lane, without any restrictions here.

[0057] Correcting the initial local map based on historical perception data, target perception data, and target precise map data to obtain the target local map can be understood as comprehensively using historical perception data, target perception data, and target precise map data to perform multi-dimensional correction and supplement on the initial local map originally generated by the target perception data to improve the accuracy and timeliness of the map.

[0058] The target local map includes at least one drivable lane, which can be understood as the target local map clearly identifying at least one lane in which the vehicle can safely pass under current conditions, thereby helping to improve driving safety and efficiency.

[0059] In this embodiment of the present invention, historical perception data provides information perceived by the vehicle in the past under similar or similar road conditions, helping to identify common obstacles or environmental features on the road. Target perception data provides real-time information about the vehicle's current surroundings, reflecting the real-time road conditions. Target precision map data provides the road's basic framework and attributes. While less accurate than high-precision maps, it ensures the correctness of the map's basic structure. By integrating these data and correcting potential errors or incomplete information in the initial local map, the target local map can more comprehensively reflect the actual road conditions on the target vehicle's route, including road topology, obstacle distribution, and drivable areas, providing a more accurate basis for lane recommendations.

[0060] Step S13: determining a target lane from at least one drivable lane based on a preset road scoring strategy, so as to display the target lane on a display interface, wherein the preset road scoring strategy is used to score any drivable road according to at least one scoring rule.

[0061] In an embodiment of the present invention, the preset road scoring strategy can be understood as a pre-designed algorithm or rule set for evaluating and comparing the suitability of all drivable lanes in the target local map. For example, the scoring rules included in the scoring strategy can be set according to different factors and goals, such as assigning a comprehensive score to each lane based on the traffic flow of the lane, the proportion of obstacles on the lane, the consistency of the lane with the target route, and the user's historical driving preferences. The purpose of the preset road scoring strategy is to convert multiple complex road conditions and driving goals into comparable numerical values through quantitative evaluation, so that the lane that best suits the current driving situation can be objectively and systematically selected, which is not limited here.

[0062] The target lane can be understood as the best lane selected from all drivable lanes in the target local map under the guidance of the preset road scoring strategy. For example, the selection of the target lane is based on the comprehensive evaluation of the lanes in the preset scoring strategy, and the lane with the highest score will usually be determined as the target lane. For example, after considering factors such as traffic flow, obstacles and consistency of the target route, a lane with moderate traffic flow, no obvious obstacles and consistent with the target route may be selected as the target lane. In the display interface, the target lane will be presented to the driver with eye-catching visual cues (such as highlights, arrow instructions, text annotations, etc.) to help them make safer and more efficient driving decisions, which are not restricted here.

[0063] The display interface can be understood as an interface for displaying the target lane. For example, the display interface can be a car navigation system screen, a head-up display, a car instrument panel, a voice prompt, a driver assistance interface, etc., which are not limited here.

[0064] Determining a target lane from at least one drivable lane based on a preset road scoring strategy and displaying the target lane on a display interface can be understood as applying the preset road scoring strategy to evaluate and compare all drivable lanes in the target local map, selecting the optimal lane, and highlighting this lane on the vehicle's display screen or other interactive interface to guide the driver or the autonomous driving system in driving decisions. For example, various previously collected and processed information, including real-time perception data, historical driving data, and accurate map data, can be utilized to comprehensively consider various lane attributes and conditions, ultimately selecting a lane that best meets current driving needs and objectives, thereby ensuring safer, smoother, and more efficient driving. This is not a limitation herein.

[0065] The preset road scoring strategy is used to score any drivable road according to at least one scoring rule. It can be understood that, based on a pre-set scoring system, the characteristics of each drivable lane are automatically analyzed, such as traffic density, obstacle conditions, lane type (for example, straight lane, turning lane), road conditions (for example, wetness, clarity of sight), etc., and a score is assigned to each lane. The high or low score reflects the quality of the lane under the current driving conditions. For example, the preset road scoring strategy usually includes scoring rules at multiple levels or dimensions to comprehensively examine different aspects of the lane. The design of the scoring rules and the distribution of weights need to take into account various factors that may affect driving safety and efficiency to ensure that the final scoring results can objectively reflect the comprehensive quality of the lane, and then accurately guide the driver or the automatic driving system to select the optimal lane. There is no restriction here.

[0066] In an embodiment of the present invention, a preset scoring strategy is used to conduct an in-depth analysis and comparison of all drivable lanes in the target local map, select a lane with the best overall conditions as the target lane, and display it through the vehicle's display interface, providing the driver or the autonomous driving system with real-time, accurate, and personalized lane recommendation services, which helps to improve driving safety, road traffic efficiency, and driving comfort.

[0067] Figure 2 is a flow chart of a lane recommendation method according to an embodiment of the present invention. Figure 2As shown in Figure 1, the lane recommendation method involves using onboard sensors (such as cameras and lidar) to collect road information in real time, including detailed information such as road shape, lane markings, traffic signs, and obstacles. A local map is created with the ego vehicle as the origin, and the pose of all collected information is expressed in the ego vehicle coordinate system. The collected data is first preprocessed and cleaned to remove noise and outliers. Key road components are then identified, such as crosswalks, guide lines, forks, intersection frames, and drivable areas. Roads are then divided into multiple sections based on intersections, and lanes are divided into multiple sections based on lane markings. Topological relationships between roads and lanes are established. Positioning technologies such as the Global Positioning System (GPS) and Real-Time Kinematic (RTK) are then used to obtain the vehicle's precise position. The information in the standard map data is converted from the world coordinate system to the ego vehicle coordinate system. The standard map data and real-time perception data are then compared and fused in the ego vehicle coordinate system. If mismatched information occurs, the system uses the information from the previous frame obtained from historical perception data and the current high-confidence standard map data to correct it and then fuses it with the real-time perception data. For example, if a lane is unavailable due to temporary construction, the perception information prevails. Ultimately, the topological relationship is corrected based on the fused information, resulting in a lane-focused local map. Unlike HD maps, where the overall position, direction, width, length, curvature of the road, as well as the specific location and boundaries of each lane, must be precisely aligned and corrected, the present invention focuses on the number of lanes, lane type, lane drivability, and the percentage of obstacles within the lanes. During navigation, the map engine displays available lanes and sometimes recommends the optimal lane based on the current route and traffic conditions. For example, this information can be flexibly displayed at the most appropriate location on devices such as the in-vehicle navigation screen, instrument panel, head-up display, smart rearview mirror, and in-vehicle infotainment system, based on real-time road conditions and the driver's specific needs. Combining the vehicle's location with navigation recommendations derived from standard map data, the system perceives real-time road conditions and assigns a multi-level score to available lanes, ultimately recommending the highest-scoring lane to the user.

[0068] Figure 3 FIG. 1 is a flow chart of a scoring strategy according to an embodiment of the present invention. Figure 3As shown, the system first obtains the total number of lanes and the ordinal numbers of the currently drivable lanes, including the number of the lane the ego vehicle is currently in. This ensures the system understands which lanes are drivable and the ego vehicle's current position. Next, the system scores all drivable lanes based on their relative distance from the ego vehicle. Lane distance is scored based on a "high for close, low for far" principle: lanes closer to the ego vehicle receive higher scores, while lanes closer to the ego vehicle receive lower scores. Non-drivable lanes receive a score of zero. This strategy reflects the driver's preference for nearby lanes and the exclusion of non-drivable lanes. The system then checks for obstacles and, if present, scores the lane based on the volume percentage of the obstacle. A higher percentage of obstacles results in a lower score, with non-drivable lanes receiving a score of zero. This step ensures the system considers real-time lane availability and prioritizes lanes with no or minimal obstacles, improving driving safety and smoothness. The system also considers lane recommendations made by the map engine during navigation, assigning additional points to recommended lanes, while non-drivable lanes receive a score of zero. This system leverages the navigation system's pre-planned routes and comprehensive consideration of road conditions, making lane recommendations more closely aligned with the driver's actual needs and target route. If the system has access to the user's historical driving data, it analyzes the user's past lane selection preferences on the same road section, awarding points to frequently selected lanes while retaining a zero score for unavailable lanes. This scoring mechanism respects user habits and provides personalized service, enabling lane recommendations to better meet individual user needs. After completing all the above scoring steps, the system combines the results of all scoring items to calculate a total score for each drivable lane. The lane with the highest total score is designated as the target lane and presented intuitively to the driver on the vehicle's display interface. This meticulous scoring strategy ensures that the lane recommendation system comprehensively considers multiple lane attributes and real-time changes in the driving environment, resulting in more intelligent, safe, and efficient lane recommendations. This effectively guides the driver or the autonomous driving system to select the optimal lane, enhancing the overall driving experience.

[0069] In this embodiment of the present invention, on-board sensors first collect real-time environmental information about the target vehicle's roadway to construct an initial local map depicting the current road topology, including lane layout and connectivity. This map is then corrected using historical perception data and target precision map data to form a target local map. This map ensures that it includes detailed information about at least one drivable lane and corrects errors caused by temporary changes. Finally, based on a pre-set road scoring strategy, all drivable lanes are scored based on multiple factors, including lane conditions, obstacle distribution, navigation guidance, and user preferences. The highest-scoring lane is selected as the target lane and intuitively displayed on the vehicle's display interface, guiding the driver in making the optimal lane selection. This reduces the user's reliance on high-precision map data, thereby achieving a low-cost, highly applicable, and real-time lane recommendation service. This addresses the technical challenge of providing accurate and real-time lane recommendation services to improve driving efficiency while reducing costs in the absence of high-precision map data.

[0070] Optionally, the target local map further includes at least one of the following: lane number information, lane type information, lane drivable information, and obstacle ratio information on the lane.

[0071] In the embodiment of the present invention, lane number information can be understood as the number of lanes identified by the system in a certain area. For example, at an intersection or a section of a highway, there may be multiple parallel lanes for vehicles to travel, which is not limited here.

[0072] Lane type information can be understood as describing the function or attributes of the lane, such as fast lane, slow lane, bus lane, merging lane, diverging lane, turning lane, etc., which is not limited here.

[0073] Lane drivability information can be understood as information reflecting whether the lane is safe to traverse at the current moment. For example, if there are obstacles in the lane (such as temporary construction, traffic accidents, severe congestion, etc.), the lane will be marked as drivable. Conversely, if the lane is unobstructed, it will be marked as drivable, and this is not restricted here.

[0074] The obstacle percentage information on a lane can be understood as information used to quantify the density of obstacles on a lane, typically expressed as the percentage of the lane area occupied by obstacles. For example, a high obstacle percentage may indicate a congested or dangerous lane, thereby lowering the lane's score in the lane recommendation scoring. This is not a limitation here.

[0075] In an embodiment of the present invention, a local map is constructed based on information about the number of lanes, lane types, whether the lanes are drivable, and the proportion of obstacles on the lanes, and the most suitable lanes are recommended. This is not only cost-effective and real-time, but also widely applicable, capable of coping with various road environments and improving driving efficiency and safety.

[0076] Optionally, in step S10, obtaining target perception data includes the following steps:

[0077] Step S101, obtaining initial perception data;

[0078] Step S102, performing data preprocessing on the initial perception data to obtain preprocessed perception data;

[0079] Step S103: perform data labeling on the pre-processed perception data to obtain target perception data.

[0080] In the embodiments of the present invention, initial perception data can be understood as raw data directly collected by the vehicle's sensors (such as cameras, radars, lidars, etc.). For example, initial perception data may include road images, depth information of the surrounding environment, obstacle detection results, etc. Because the sensors output unprocessed information, initial perception data may include a large amount of background noise, redundant information, and unclassified objects, which are not limited here.

[0081] Data preprocessing can be understood as a series of cleaning, filtering, and formatting operations performed by the system on the initial sensor data. For example, the goals of data preprocessing are to remove noise, eliminate redundant information, correct for sensor bias, and convert the data into a form suitable for subsequent processing, although this is not a limitation.

[0082] Preprocessed perception data can be understood as clean, standardized perception data that has undergone data preprocessing. For example, preprocessed perception data removes most noise from the original data, highlighting key information directly relevant to lane recommendations, such as road characteristics and obstacle locations. This is not a limitation here.

[0083] Data annotation can be understood as the process of manually or automatically labeling objects in the pre-processed sensory data. For example, in a lane recommendation system, data annotation may include identifying and marking lane lines, obstacles, traffic signs, drivable areas, etc., but this is not limited here.

[0084] Obtaining initial perception data can be understood as collecting the original data of the vehicle's surrounding environment obtained by sensors on the vehicle (such as cameras, lidar, millimeter-wave radar, etc.), which may include images captured by visual sensors, point cloud data generated by lidar, echo data of millimeter-wave radar, etc., which are not restricted here.

[0085] Performing data preprocessing on the initial perception data to obtain preprocessed perception data can be understood as, after obtaining the original sensor data, the system will remove useless or interfering information, such as noise, background, inherent bias of the sensor, etc., and convert the data into a format that the algorithm can process more efficiently, thereby obtaining preprocessed perception data, which is not limited here.

[0086] Data labeling of the preprocessed perception data to obtain target perception data can be understood as follows: after data preprocessing, the system will label the preprocessed data, that is, add labels or classification information to key elements in the image or point cloud data (such as lane lines, vehicles, pedestrians, road signs, etc.), thereby obtaining target perception data, which is not restricted here.

[0087] In an embodiment of the present invention, raw data of the vehicle's surrounding environment is first collected through the vehicle's sensors, such as cameras and radars, to form initial perception data. Subsequently, data preprocessing is performed on this initial perception data, including noise removal, sensor bias calibration, data format conversion, etc., to improve data quality, thereby obtaining clearer and more applicable preprocessed perception data. Finally, the preprocessed perception data is finely labeled, and classification labels are added to key elements in the image, such as lane lines, obstacles, and traffic signs, to convert them into target perception data, thereby providing cleaned, labeled, and optimized data for the subsequent lane recommendation algorithm, ensuring that the system can make decisions and operations based on accurate and efficient information.

[0088] Optionally, in step S11, generating an initial local map based on target perception data includes the following steps:

[0089] Step S111, determining the location information of the target vehicle;

[0090] Step S112, dividing multiple roads based on the intersection data in the target perception data;

[0091] Step S113, dividing a plurality of lanes based on lane line data in the target perception data;

[0092] Step S114: generating an initial local map based on the location information, the multiple roads, and the multiple lanes.

[0093] In the embodiments of the present invention, the target vehicle's location information can be understood as the vehicle's current precise geographic location data, including but not limited to latitude and longitude coordinates, altitude, speed, heading, and other information. For example, the target vehicle's location information is typically acquired by an onboard positioning system (such as GPS, RTK, etc.) to determine the vehicle's precise location on the road, which is not limited herein.

[0094] Intersection data can be understood as intersection-related information captured in real time by on-board sensors (such as cameras and radars), including the shape, size, directional arrows, traffic lights, stop lines, lane distribution, etc. of the intersection, which is not restricted here.

[0095] Lane data can be understood as lane dividing line information on the road detected by the vehicle's sensors. For example, lane data typically includes lane line type (solid line, dashed line, double yellow line, etc.), position, direction, and continuity, which are not limited here.

[0096] Determining the location information of the target vehicle can be understood as obtaining the vehicle's current precise location and posture information on the earth's surface through the vehicle-mounted positioning system, including the vehicle's latitude and longitude coordinates, altitude, heading angle, driving speed, etc., which are not restricted here.

[0097] Dividing multiple roads based on intersection data in target perception data can be understood as using real-time environmental data collected by vehicle sensors (such as cameras and lidar), especially information about the intersection area, including the shape, size, guide arrows, lane distribution, etc. of the intersection, so that the system can identify and divide multiple roads, which is not restricted here.

[0098] Dividing multiple lanes based on lane line data in target perception data can be understood as the system analyzing the lane line images or point cloud data obtained by the sensor to identify the type of lane line (solid line, dotted line, etc.), position and direction, so that multiple lanes can be divided. This is not restricted here.

[0099] Generating an initial local map based on location information, multiple roads, and multiple lanes can be understood as integrating the vehicle's real-time location information, the identified multiple roads and the topological relationships between them, and the division results of multiple lanes. The system can construct a local map centered on the vehicle, which is not restricted here.

[0100] In an embodiment of the present invention, first, the system determines the precise location information of the target vehicle on the earth's surface, including position, heading, and speed, through the on-board positioning equipment. Then, based on the environmental data in the target area perceived by the vehicle sensors, the system identifies and divides multiple roads and lanes, paying special attention to intersection and lane line data, to construct a detailed road structure and lane distribution. Finally, combining the real-time location of the vehicle and the information of the multiple divided roads and lanes, the system generates an initial local map centered on the target vehicle. The initial local map contains a detailed description of the vehicle's surrounding environment, such as the number, type, traffic status, and connection relationship of lanes with other roads, providing real-time environmental awareness for intelligent vehicles and ensuring driving safety and efficiency.

[0101] Optionally, in step S10, obtaining target map data includes the following steps:

[0102] Step S104, obtaining initial precise map data, wherein the initial precise map data is map data in a world coordinate system;

[0103] In step S105 , the initial accurate map data is converted to the vehicle coordinate system to obtain the target accurate map data.

[0104] In an embodiment of the present invention, the initial precise map data can be understood as map data that contains road network information but has lower accuracy than high-precision maps (usually with an accuracy of several meters to more than ten meters). For example, the initial precise map data is sufficient to describe the basic topological structure of the road, including the location, direction, number of lanes, lane type, etc. of the road, but is not sufficient to support the high-precision positioning and environmental perception required for autonomous driving. The initial precise map data is represented in the world coordinate system, that is, each element on the map (such as roads, buildings, landmarks, etc.) has its exact position in the global coordinate system, which is not limited here.

[0105] The world coordinate system can be understood as an absolute coordinate system used for precise positioning on the surface of the earth. For example, in maps and geographic information systems, the world coordinate system is usually based on the longitude and latitude coordinates of the earth, or uses a projection to convert the earth's surface into a two-dimensional plane coordinate system, which is not limited here.

[0106] The ego-vehicle coordinate system can be understood as a coordinate system relative to the vehicle's own position. For example, in intelligent vehicle technology, the ego-vehicle coordinate system is often used to describe the relative position of the vehicle's surroundings. For example, the position of an obstacle 10 meters in front of the vehicle in the ego-vehicle coordinate system is closely related to the vehicle's direction of travel and current position, and is not limited here.

[0107] Target precise map data can be understood as the data obtained by converting the initial precise map data from the world coordinate system to the vehicle coordinate system. For example, target precise map data includes information such as the number, type, and drivability of lanes on the roads surrounding the vehicle, which helps the vehicle understand and process road information from its own perspective. This is not limited here.

[0108] Converting the initial precision map data to the vehicle coordinate system to obtain the target precision map data can be understood as converting the initial precision map data from a global, static perspective to a dynamic perspective that matches the vehicle's current position and orientation, thereby obtaining the target precision map data. For example, the vehicle's real-time positioning data (such as GPS coordinates, heading angle, etc.) is used to recalculate the position of each point in the map data relative to the vehicle. Through coordinate conversion, information such as roads, lanes, and traffic signs originally in the world coordinate system are now represented as directions and distances relative to the vehicle's current position, thereby making the map data more consistent with the vehicle's real-time driving environment and facilitating the system's decision-making based on the vehicle's dynamic state. This is not limited here.

[0109] In this embodiment of the present invention, the system first acquires initial precision map data, which describes the basic structure of the road network in a world coordinate system, including road locations, directions, and lane information. Subsequently, using the vehicle's real-time positioning information, this global coordinate system map data is converted to a vehicle coordinate system with the vehicle itself as the origin, forming target precision map data. This ensures that the map data can be dynamically adjusted based on the vehicle's real-time position and direction, making it more accurate to the vehicle's actual driving environment.

[0110] Optionally, in step S12, correcting the initial local map based on the historical perception data, the target perception data, and the target precise map data to obtain the target local map includes the following steps:

[0111] Step S121, matching the target perception data with the target map data to obtain an initial matching result;

[0112] Step S122, in response to the matching result including a matching failure message, determining the previous perception data of the previous image frame based on the historical perception data, wherein the timestamp corresponding to the previous image frame is earlier than the timestamp corresponding to the current image frame, and the target perception data is the perception data of the current image frame;

[0113] Step S123: Correct the matching failure information based on the confidence level of the previous perception data and the confidence level of the target map data to obtain matching success information;

[0114] Step S124, updating the initial matching result based on the matching success information to obtain the target matching result;

[0115] Step S125 , correcting the initial local map based on the target matching result to obtain a target local map.

[0116] In the embodiment of the present invention, the initial matching result can be understood as the result of preliminary matching between the vehicle's target perception data (such as real-time environmental information obtained by cameras, radars, etc.) and the standard map data.

[0117] Match failure information can be understood as information indicating a discrepancy between the current target perception data and the standard map data. For example, during the matching process, if there is an inconsistency or mismatch between the target perception data and the standard map data, such as a road or lane change caused by temporary construction, the system will generate a match failure message, which is not limited here.

[0118] The previous perception data of a previous image frame can be understood as the environmental information captured by the vehicle's sensors in the previous image frame. For example, the previous perception data of the previous image frame corresponds to a time earlier than the current frame, but may contain a more accurate or continuous description of the current environment, which is not limited here.

[0119] The confidence level of the preceding perception data can be understood as the reliability of the preceding perception data. For example, the preceding perception data with a high confidence level will be given priority when correcting the matching failure information, which is not limited here.

[0120] The confidence level of the target precise map data can be understood as the reliability of the target precise map data. For example, precise map data with a high confidence level will also be given more weight when being revised, which is not limited here.

[0121] The successful matching information can be understood as the system being able to obtain the precise map data information that is consistent with the current perception environment after correcting the failed matching information.

[0122] The target matching result can be understood as the final matching result obtained after correcting and integrating the previous perception data and the standard map data. The target matching result integrates the optimal information of the previous perception data and the standard map data, and improves the accuracy of the local map. It is not restricted here.

[0123] Matching the target perception data and the target precise map data to obtain the initial matching result can be understood as comparing and fusing the current environmental information (including roads, lane lines, obstacles, etc.) captured by the vehicle's real-time perception system (through sensors such as cameras and radars) with the target precise map data (including basic road layout and attribute information) to confirm the consistency and correspondence between the two, which is not restricted here.

[0124] In response to the matching result including a match failure message, determining the preceding perception data of the preceding image frame based on the historical perception data can be understood as follows: when the system detects that there is a mismatch or a match failure between the current perception data and the precision map data, the system will backtrack to the perception data of the previous moment (or previous frames), using historical information to assist in explaining the current inconsistency and finding a more reasonable data explanation or correction basis. For example, the preceding perception data is earlier in time than the current frame data, providing a continuity perspective that helps understand and correct the current perception deviation, which is not limited here.

[0125] The timestamp corresponding to the previous image frame is earlier than the timestamp corresponding to the current image frame. The target perception data is the perception data of the current image frame. This means that when processing the real-time perception data of the current image frame, the system will refer to the image frame data corresponding to the previous timestamp, i.e., the previous perception data, to enhance the accuracy and interpretability of the current perception data by leveraging the continuity and consistency of historical data. The target perception data is the latest environmental perception information currently being processed by the system and is not limited here.

[0126] Correcting match failure information based on the confidence of the previous perception data and the confidence of the target precision map data to obtain successful match information can be understood as utilizing the confidence information of historical perception data (previous perception data) and precision map data to analyze and correct match failures to improve matching accuracy. For example, data sources with higher confidence levels will have a greater impact on the correction results. Through comparison and fusion, the system can convert original failed match information into successful match information, which is not limited here.

[0127] Updating the initial matching result based on the successful matching information to obtain the target matching result can be understood as integrating the corrected successful matching information into the existing initial matching result, updating and optimizing any previously erroneous or incomplete matching information. For example, the target matching result is the latest matching state that integrates real-time perception data and accurate map data, reflecting the system's most accurate understanding of the current environment. This is not a limitation here.

[0128] Modifying the initial local map based on the target matching results to obtain a target local map can be understood as modifying the previously constructed initial local map based on the target matching results, updating information such as roads, lanes, and obstacles to better reflect the current real-world environment. For example, the target local map is a modified and optimized view of the local environment, providing the vehicle with more accurate navigation and decision-making. This is not a limitation here.

[0129] In an embodiment of the present invention, the system first matches the currently perceived environmental information with the standard map data to obtain an initial matching result. When matching failure information is found, the system uses historical perception data (previous perception data) to correct this failure information. Based on the confidence level of the previous perception data and the standard map data, it determines which information is more reliable and updates the initial matching result accordingly, ultimately obtaining a target matching result. Based on the target matching result, the system makes corresponding corrections to the initial local map to make it more closely match the current real environment, generating a target local map, providing the vehicle with accurate lane-level navigation information, and supporting safer and more efficient driving decisions.

[0130] Optionally, in step S13, determining a target lane from at least one drivable lane based on a preset road scoring strategy includes the following steps:

[0131] Step S131, determining navigation recommended lane information based on the target map data;

[0132] Step S132: scoring at least one drivable lane based on a preset road scoring strategy, navigation-recommended lane information, and target perception data to obtain a lane score value;

[0133] Step S133: determining a target lane from at least one drivable lane based on the lane scoring value.

[0134] In an embodiment of the present invention, the navigation recommended lane information can be understood as referring to the suggested or recommended lane information provided by the navigation system based on the target map data, which is used to guide the vehicle to select the most suitable lane for driving in the road network. Exemplarily, the navigation recommended lane information may include the following content: the recommended lane sequence number or lane number, which is used to indicate which lane the vehicle should drive on. Lane type, such as ordinary lane, bus lane, turning lane or straight lane, etc. The traffic status of the lane, such as whether it is unobstructed, whether there is temporary construction, whether it is a priority lane, etc. The association between the lane and the vehicle's destination, which is used to indicate whether the lane leads directly to the destination or whether it is the best choice to reach the destination, is not limited here.

[0135] Determining recommended navigation lane information based on target precise map data can be understood as the system utilizing precise map data, including information about road structure, lane distribution, traffic rules, and road attributes, combined with the vehicle's current location and destination, to calculate recommended lane information through a navigation algorithm. For example, recommended navigation lane information may include suggestions to maintain the vehicle's current lane, change to a specific lane, or prepare to enter an exit lane in advance. This information is intended to guide the vehicle along the optimal path, avoid traffic violations or incorrect routes, and improve driving efficiency and safety. This is not a limitation here.

[0136] Based on a preset road scoring strategy, navigation-recommended lane information, and target perception data, the system scores at least one drivable lane. This results in a lane score. This can be understood as the system applying a pre-set road scoring strategy, comprehensively considering the navigation-recommended lane information and the vehicle's current real-time road condition information (i.e., target perception data) collected by sensors, to score all drivable lanes. For example, the scoring strategy may include multiple dimensions, such as lane congestion, the presence of obstacles, proximity to the destination route, and compliance with traffic regulations. This quantitative evaluation of each lane's suitability yields a score, which is not limited here.

[0137] Determining a target lane from at least one drivable lane based on a lane score can be understood as the system selecting the lane with the highest score or that meets specific conditions (such as shortest driving distance, shortest driving time, and lowest fuel consumption) as the target lane based on the evaluation and scoring results (i.e., lane scores) of each drivable lane. For example, the target lane is the lane that the system determines best meets current driving requirements and safety standards and is recommended to the driver or the autonomous driving system to guide the vehicle's driving decisions. This is not a limitation here.

[0138] In this embodiment of the present invention, the system first determines navigation-recommended lane information based on the target map data. Next, using a pre-set scoring strategy combined with real-time road conditions (i.e., target perception data) and the recommended lane information, it evaluates the vehicle's currently drivable lanes and generates a score for each lane. Finally, based on the lane scores, the system selects the optimal lane (i.e., the target lane) from all drivable lanes, providing the most appropriate driving recommendations for the vehicle, ensuring an efficient and safe driving experience.

[0139] Optionally, scoring at least one drivable lane based on a preset road scoring strategy, navigation recommended lane information, and target perception data to obtain a lane score value includes the following steps:

[0140] Based on the target perception data, the distance between any drivable lane and the target vehicle, the volume ratio of obstacles on any drivable lane, and the historical driving data of any drivable lane are obtained;

[0141] Determine a first score for any drivable lane based on distance, determine a second score for any drivable lane based on volume percentage, determine a third score for any drivable lane based on navigation-recommended lane information, and determine a fourth score for any drivable lane based on historical driving data;

[0142] A lane score value of any drivable lane is determined based on the first score value, the second score value, the third score value, and the fourth score value.

[0143] In this embodiment of the present invention, the first score can be understood as a score assigned to a lane based on the distance between any drivable lane and the target vehicle. For example, lanes closer to the target vehicle may receive a higher first score, while lanes farther away may receive a lower first score, although this is not a limitation.

[0144] The second score can be understood as a score assigned to any drivable lane based on the proportion of the lane space occupied by obstacles. For example, if the volume of obstacles in a lane is large, the second score is low, indicating poor traffic efficiency and safety in that lane. Conversely, lanes with a small volume of obstacles or even no obstacles have a higher second score, indicating greater suitability for vehicle travel, and this is not limited here.

[0145] The third score can be understood as a score assigned to a lane based on how closely it matches the vehicle's navigation path. For example, if a lane is recommended by the navigation system, its third score is higher, indicating that it is more likely to be the optimal path to the destination. For non-recommended lanes, the third score is lower, and this is not a limitation here.

[0146] The fourth score can be understood as a lane score based on historical statistics such as lane usage frequency, average speed, and accident history under similar conditions. If a lane has been frequently used and has a high average speed, its fourth score may be high, indicating that under similar conditions, it offers good efficiency and comfort. Conversely, lanes with frequent accidents or slow traffic may have lower fourth scores and are not subject to this restriction.

[0147] Obtaining the distance between any drivable lane and the target vehicle, the volume ratio of obstacles on any drivable lane, and the historical driving data of any drivable lane based on target perception data can be understood as the system using real-time environmental perception information (i.e., target perception data) to obtain key driving parameters, including but not limited to: the relative position relationship between the current drivable lane and the vehicle, i.e., the lane distance; the proportion of space occupied by obstacles on the lane, where the volume ratio is used to quantify the impact of obstacles on the lane's traffic capacity; and lane historical data obtained by analyzing records of past vehicles driving under the same or similar conditions, such as average driving speed, frequency of use and other indicators, which are not restricted here.

[0148] The first score for any drivable lane is determined based on distance, the second score is determined based on volume percentage, the third score is determined based on navigation-recommended lane information, and the fourth score is determined based on historical driving data. This means the system uses a multi-level scoring mechanism to evaluate each drivable lane. For example, the distance between the lane and the vehicle is converted into a first score, with closer distances likely resulting in higher scores. The volume percentage of the obstacle determines the second score, with lower percentages likely resulting in higher scores, reflecting the lane's capacity. Lanes recommended by the navigation system receive a higher third score, reflecting the lane's relevance to the destination. Historical driving data, such as the lane's past efficiency and safety record, is used to determine the fourth score, with higher scores representing better historical performance, which is not limited here.

[0149] Determining the lane score for any drivable lane based on the first, second, third, and fourth scores can be understood as the system comprehensively considering these four scores and converting them into a comprehensive evaluation metric, namely the lane score, using a weighted average, weighted sum, or other method. For example, the lane score reflects an overall assessment of the lane's suitability for the target vehicle, taking into account distance, obstacle impact, navigation recommendations, and historical driving data. A higher score indicates a more suitable lane for the current vehicle's driving needs, and can serve as an important basis for lane recommendation services, without limitation here.

[0150] In an embodiment of the present invention, the system first uses target perception data to analyze the relative distance between any drivable lane and the target vehicle in real time, assess the volume percentage of obstacles on the lane, and retrieve historical driving data for the lane, including but not limited to the lane's average speed and frequency of use. Based on this detailed analysis, the system then calculates a first score, a second score, a third score, and a fourth score based on four dimensions: physical distance, obstacles, navigation guidance, and historical performance, to quantitatively assess the lane's suitability. Finally, the system comprehensively considers these four scores and uses a specific algorithm to determine the final lane score for each lane. Lanes with higher scores are considered more optimal driving options. This allows the system to intelligently recommend at least one drivable lane in the absence of high-precision map data by integrating real-time perception, positioning technology, and standard map data, aiming to improve vehicle safety, efficiency, and driving experience.

[0151] The present invention proposes a lane determination method that does not rely on high-cost and complex-to-maintain high-precision map data. Instead, it integrates the basic lane information of the standard high-precision map with the real-time perception data of the vehicle, thereby realizing intelligent and real-time lane recommendation. The present invention not only greatly broadens the application scope of the lane recommendation service, making it effective on highways, urban roads and even rural roads, but also significantly reduces the overall cost of the service, because lane-level navigation guidance can be completed only by relying on on-board sensors and relatively simple standard high-precision maps. More importantly, with the help of real-time environmental perception, the present invention can quickly respond to road changes, such as traffic condition updates, thereby effectively avoiding safety hazards caused by outdated map information. In addition, the intelligent design of the recommendation strategy comprehensively considers the vehicle's position, obstacle distribution, user preferences and navigation suggestions, ensuring the comprehensiveness and accuracy of the recommendation results, greatly improving the driving experience and road traffic efficiency.

[0152] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0153] According to an embodiment of the present invention, an embodiment of a lane determination device is provided. It should be noted that the device can be used to execute the above lane determination method.

[0154] Figure 4 is a module diagram of a lane determination device according to an embodiment of the present invention. Figure 4 As shown, the lane determination device 400 includes: an acquisition module 401, used to obtain target landmark map data and target perception data, wherein the target perception data is used to represent the road information of the road where the target vehicle is located; a generation module 402, used to generate an initial local map based on the target perception data, wherein the initial local map is used to describe the road topology relationship of the road where the target vehicle is located; a correction module 403, used to correct the initial local map based on historical perception data, target perception data and target landmark map data to obtain a target local map, wherein the target local map includes at least one drivable lane; a determination module 404, used to determine the target lane from at least one drivable lane based on a preset road scoring strategy to display the target lane in a display interface, wherein the preset road scoring strategy is used to score any drivable road according to at least one scoring rule.

[0155] Optionally, the target local map further includes at least one of the following: lane number information, lane type information, lane drivable information, and obstacle ratio information on the lane.

[0156] Optionally, the acquisition module 401 is further configured to acquire initial perception data; perform data preprocessing on the initial perception data to obtain preprocessed perception data; and perform data labeling on the preprocessed perception data to obtain target perception data.

[0157] Optionally, the generation module 402 is also used to determine the location information of the target vehicle; divide multiple roads based on intersection data in the target perception data; divide multiple lanes based on lane line data in the target perception data; and generate an initial local map based on the location information, multiple roads, and multiple lanes.

[0158] Optionally, the acquisition module 401 is further configured to acquire initial precise map data, wherein the initial precise map data is map data in a world coordinate system; and convert the initial precise map data into a vehicle coordinate system to obtain target precise map data.

[0159] Optionally, the correction module 403 is also used to match the target perception data and the target precision map data to obtain an initial matching result; in response to the matching result including matching failure information, determine the previous perception data of the previous image frame based on the historical perception data, wherein the timestamp corresponding to the previous image frame is earlier than the timestamp corresponding to the current image frame, and the target perception data is the perception data of the current image frame; based on the confidence of the previous perception data and the confidence of the target precision map data, the matching failure information is corrected to obtain matching success information; based on the matching success information, the initial matching result is updated to obtain the target matching result; based on the target matching result, the initial local map is corrected to obtain the target local map.

[0160] Optionally, the determination module 404 is also used to determine the navigation recommended lane information based on the target map data; score at least one drivable lane based on a preset road scoring strategy, the navigation recommended lane information and the target perception data to obtain a lane scoring value; and determine the target lane from at least one drivable lane based on the lane scoring value.

[0161] Optionally, the determination module 404 is also used to obtain the distance between any drivable lane and the target vehicle, the volume ratio of obstacles on any drivable lane, and the historical driving data of any drivable lane based on the target perception data; determine the first score of any drivable lane based on the distance, determine the second score of any drivable lane based on the volume ratio, determine the third score of any drivable lane based on the navigation recommended lane information, and determine the fourth score of any drivable lane based on the historical driving data; determine the lane score value of any drivable lane based on the first score, the second score, the third score and the fourth score.

[0162] According to another aspect of an embodiment of the present invention, a vehicle is further provided, comprising a computer program, which implements the methods in various embodiments of the present invention when executed by a processor.

[0163] According to another aspect of an embodiment of the present invention, an electronic device is provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods of various embodiments of the present invention when running.

[0164] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the methods in various embodiments of the present invention.

[0165] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the method in each embodiment of the present invention is implemented.

[0166] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0167] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only exemplary. For example, the division of the units can be a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0168] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0169] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0170] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program codes.

[0171] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A lane determination method, characterized in that: include: Obtaining target map data and target perception data, wherein the target perception data is used to represent road information of the road where the target vehicle is located; generating an initial local map based on the target perception data, wherein the initial local map is used to describe a road topology relationship of a road where the target vehicle is located; Correcting the initial local map based on the historical perception data, the target perception data, and the target precise map data to obtain a target local map, wherein the target local map includes at least one drivable lane; A target lane is determined from the at least one drivable lane based on a preset road scoring strategy to display the target lane in a display interface, wherein the preset road scoring strategy is used to score any drivable road according to at least one scoring rule.

2. The method according to claim 1, characterized in that The target local map further includes at least one of the following: lane number information, lane type information, lane drivable information, and obstacle ratio information on the lane.

3. The method according to claim 1, characterized in that Acquiring the target perception data includes: Acquire initial perception data; Performing data preprocessing on the initial perception data to obtain preprocessed perception data; Data annotation is performed on the preprocessed perception data to obtain the target perception data.

4. The method according to claim 3, characterized in that Generating an initial local map based on the target perception data includes: Determining the location information of the target vehicle; Dividing a plurality of roads based on intersection data in the target perception data; Dividing a plurality of lanes based on lane line data in the target perception data; The initial local map is generated based on the position information, the plurality of roads, and the plurality of lanes.

5. The method according to claim 1, wherein Acquiring the target map data includes: Acquire initial precise map data, wherein the initial precise map data is map data in a world coordinate system; The initial accurate map data is converted to the vehicle coordinate system to obtain the target accurate map data.

6. The method according to claim 5, characterized in that The step of correcting the initial local map based on the historical perception data, the target perception data, and the target precise map data to obtain the target local map includes: Matching the target perception data with the target precise map data to obtain an initial matching result; In response to the matching result including a match failure message, determining, based on the historical perception data, previous perception data of a previous image frame, wherein a timestamp corresponding to the previous image frame is earlier than a timestamp corresponding to the current image frame, and the target perception data is the perception data of the current image frame; Correcting the matching failure information based on the confidence of the previous perception data and the confidence of the target map data to obtain matching success information; Update the initial matching result based on the matching success information to obtain a target matching result; The initial local map is modified based on the target matching result to obtain the target local map.

7. The method according to claim 1, characterized in that Determining the target lane from the at least one drivable lane based on the preset road scoring strategy includes: Determining navigation recommended lane information based on the target map data; Scoring the at least one drivable lane based on the preset road scoring strategy, the navigation recommended lane information, and the target perception data to obtain a lane scoring value; The target lane is determined from the at least one drivable lane based on the lane score value.

8. The method according to claim 7, characterized in that Scoring the at least one drivable lane based on the preset road scoring strategy, the navigation recommended lane information, and the target perception data to obtain a lane scoring value includes: Obtaining, based on the target perception data, a distance between any drivable lane and the target vehicle, a volume ratio of obstacles on any drivable lane, and historical driving data of any drivable lane; determining a first score for any drivable lane based on the distance, determining a second score for any drivable lane based on the volume share, determining a third score for any drivable lane based on the navigation-recommended lane information, and determining a fourth score for any drivable lane based on the historical driving data; A lane score value of any drivable lane is determined based on the first score, the second score, the third score, and the fourth score.

9. A lane determination device, characterized in that: include: An acquisition module, configured to acquire target map data and target perception data, wherein the target perception data is used to represent road information of the road where the target vehicle is located; A generating module, configured to generate an initial local map based on the target perception data, wherein the initial local map is used to describe a road topology relationship of a road where the target vehicle is located; a correction module, configured to correct the initial local map based on historical perception data, the target perception data, and the target precise map data to obtain a target local map, wherein the target local map includes at least one drivable lane; A determination module is configured to determine a target lane from the at least one drivable lane based on a preset road scoring strategy, so as to display the target lane in a display interface, wherein the preset road scoring strategy is configured to score any drivable road according to at least one scoring rule.

10. A vehicle, characterized in that: The vehicle is used to execute the lane determination method described in any one of claims 1 to 8.

11. An electronic device comprising a memory and a processor, characterized in that: The memory stores a computer program, and the processor is configured to run the computer program to execute the lane determination method according to any one of claims 1 to 8.

Images