Road marking alignment method and apparatus, computer device, and storage medium

Abstract

The application relates to a road marking alignment method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring a two-dimensional road network map; performing plane feature extraction on each road section unit in the two-dimensional road network map; predicting road marking arrangement data of each road section unit according to the extracted plane features of each road section unit and a pre-trained road marking alignment model, wherein the road marking alignment model is a model representing the correlation between the plane features of each road section unit and the road marking arrangement data; and aligning road markings in each road section unit according to the predicted road marking arrangement data of each road section unit when converting the two-dimensional road network map into a three-dimensional road network map. The method can improve the display effect of road markings in each road section unit of the three-dimensional road network map.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a road marking alignment method and device, computer equipment and storage medium. BACKGROUND

[0002] With the development of computer technology, intelligent driving is applied in different business scenarios, however, various road conditions emerge in an endless stream, therefore, a more intuitive and visual three-dimensional road network map is needed.

[0003] However, various markings, such as road dashed lines, etc., are often used to guide the vehicle to travel in the road, and the two-dimensional map data rarely contains data related to road markings, therefore, when creating a three-dimensional map, the arrangement of these road markings will appear to be incoherent and unable to be aligned, resulting in poor map display effect. SUMMARY

[0004] Therefore, it is necessary to provide a road marking alignment method, device, computer equipment and storage medium capable of improving the arrangement effect of road markings in a three-dimensional road network map to solve the above technical problems.

[0005] A road marking alignment method, the method comprising: obtaining a two-dimensional road network map; performing plane feature extraction on each road segment unit in the two-dimensional road network map; predicting road marking arrangement data of each road segment unit according to the extracted plane feature of each road segment unit and a pre-trained road marking alignment model, the road marking alignment model being a model representing the correlation between the plane feature of each road segment unit and the road marking arrangement data; and aligning road markings in each road segment unit according to the predicted road marking arrangement data of each road segment unit when converting the two-dimensional road network map into a three-dimensional road network map.

[0006] In one embodiment, the road marking arrangement data comprises at least one of a marking offset value, a marking offset range value, a marking scaling ratio value and a marking scaling ratio range value.

[0007] In one embodiment, the training method of the road marking alignment model comprises: obtaining a sample two-dimensional road network map; performing plane feature extraction on each sample road segment unit in the sample two-dimensional road network map; obtaining actual road condition data of each sample road segment unit in the sample two-dimensional road network map; and training the road marking alignment model according to the plane feature and the actual road condition data of each sample road segment unit.

[0008] In one embodiment, the actual road condition data comprises at least one of a position relative relationship between each sample road segment unit, a road type of each sample road segment unit and a marking type of each sample road segment unit.

[0009] In one embodiment, the road marking alignment model is trained based on the planar features of each sample road segment unit and actual road condition data, including: generating current marking layout data for each sample road segment unit according to a preset value range; scoring each current marking layout data according to actual road condition data; and training the road marking alignment model based on the scoring results.

[0010] In one embodiment, training the road marking alignment model based on the scoring results includes: associating each sample road segment unit whose score in the current marking layout data is greater than a preset threshold with its current marking layout data; adjusting the current marking layout data of each sample road segment unit whose score in the current marking layout data is less than the preset threshold, and using the adjusted data as the current marking layout data, and proceeding to the step of scoring each current marking layout data based on actual road condition data.

[0011] In one embodiment, the method further includes: displaying a three-dimensional road network map on a display interface; and, in response to an adjustment operation on the marking layout data of a specified road segment unit in the three-dimensional road network map, returning the adjusted marking layout data to a road marking alignment model for model optimization.

[0012] A road marking alignment device, the device comprising:

[0013] The 2D map acquisition module is used to acquire 2D road network maps;

[0014] The planar feature extraction module is used to extract planar features from each road segment unit in a two-dimensional road network map;

[0015] The layout data prediction module is used to predict the layout data of each road segment unit based on the extracted planar features of each road segment unit and the pre-trained road marking alignment model. The road marking alignment model is a model that represents the relationship between the planar features of each road segment unit and the road marking layout data.

[0016] The road marking alignment module is used to align the road markings in each road segment unit according to the predicted marking layout data when converting a two-dimensional road network map into a three-dimensional road network map.

[0017] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the road marking alignment method described above.

[0018] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the road marking alignment method described above.

[0019] The aforementioned road marking alignment method, apparatus, computer equipment, and storage medium extract features from the data of a two-dimensional road network map and obtain the marking layout data corresponding to each road segment unit based on a pre-trained road marking alignment model. Then, based on the obtained marking layout data, the road markings in each road segment unit of the three-dimensional road network map are aligned when converting the two-dimensional road network map into a three-dimensional road network map, thereby improving the display effect of road markings in the three-dimensional road network map. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating a road marking alignment method in one embodiment;

[0021] Figure 2 This is a flowchart illustrating the steps involved in training a road marking alignment model in one embodiment.

[0022] Figure 3 This is a structural block diagram of a road marking alignment device in one embodiment;

[0023] Figure 4 This is a structural block diagram of the model training module for a road marking alignment device in an application example.

[0024] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0026] In one embodiment, the road marking alignment method provided in this application can be applied to a terminal. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, vehicle terminals, terminal servers, and portable wearable devices. Specifically, the terminal acquires a two-dimensional road network map; extracts planar features from each road segment unit in the two-dimensional road network map; predicts the road marking layout data of each road segment unit based on the extracted planar features and a pre-trained road marking alignment model, where the road marking alignment model is a model representing the correlation between the planar features of each road segment unit and the road marking layout data; and when converting the two-dimensional road network map into a three-dimensional road network map, aligns the road markings in each road segment unit according to the predicted road marking layout data.

[0027] In one embodiment, such as Figure 1 As shown, a road marking alignment method is provided. Taking the application of this method to a terminal as an example, the method includes the following steps:

[0028] Step S102: Obtain a two-dimensional road network map.

[0029] A two-dimensional road network map refers to a map that includes two-dimensional planar road network data. A road network is a network of roads formed by the intersection of multiple roads. A road can be divided into at least one segment unit, and a two-dimensional road network map can include two-dimensional data of at least one segment unit of at least one road. Roads here include, but are not limited to, urban roads, highways, ring roads, roundabouts, and viaducts.

[0030] Specifically, the terminal can acquire data from a two-dimensional road network map corresponding to the road network within a specified range in real time or periodically.

[0031] Step S104: Extract planar features from each road segment unit in the two-dimensional road network map.

[0032] Among them, planar features refer to the two-dimensional planar features identified from two-dimensional road network map data, that is, two-dimensional feature data that can characterize the attributes of each road segment unit. For example, it can include the relative positional relationship between each road segment unit, the road type data of each road segment unit, and the start and end position data of each road segment unit.

[0033] Specifically, the terminal uses the acquired two-dimensional road network map as the feature extraction unit, and performs feature extraction based on the two-dimensional planar data of each road segment unit to obtain the planar features corresponding to each road segment unit.

[0034] Step S106: Predict the road marking layout data of each road segment unit based on the extracted planar features of each road segment unit and the pre-trained road marking alignment model. The road marking alignment model is a model that characterizes the relationship between the planar features of each road segment unit and the road marking layout data.

[0035] Road markings refer to lane lines used to guide vehicles, and may include dashed lines, solid lines, road boundary lines, and turning or straight-ahead indicators. A road marking alignment model is a pre-trained predictive model used to output road marking layout data. This model characterizes the relationship between the planar features of each road segment unit in a two-dimensional road network map and the road marking layout data of that unit. Road marking layout data refers to the data representing how road markings are arranged between each road segment unit and its connected road segment units.

[0036] For example, the paving mark layout data may include at least one of the following: paving mark offset value, paving mark offset range value, paving mark scaling ratio value, and paving mark scaling ratio range value. The paving mark offset value or paving mark offset range value is a value representing the amount of offset of the paving mark relative to a preset baseline; the paving mark scaling ratio value or paving mark scaling ratio range value is a value representing the scaling ratio of the paving mark relative to a preset reference size.

[0037] Specifically, the terminal can call a pre-trained road marking alignment model from memory or a database, and use the extracted planar features of each road segment unit as the input parameters of the road marking alignment model to obtain the output parameters predicted and output by the road marking alignment model. That is, the output parameters are the marking layout data corresponding to each road segment unit.

[0038] Step S108: When converting the two-dimensional road network map into a three-dimensional road network map, the road markings in each road segment unit are aligned according to the predicted marking layout data of each road segment unit.

[0039] Among them, a three-dimensional road network map refers to a three-dimensional visualization map that includes three-dimensional spatial road network data.

[0040] Specifically, after the terminal obtains the road marking layout data of each road segment unit predicted and output based on the road marking alignment model, it performs a 3D road network map conversion based on the planar features of the 2D road network map and the road marking layout data corresponding to each road segment unit. During the conversion process, the corresponding types of road markings in every two adjacent road segment units are aligned according to the road marking layout data corresponding to each road segment unit.

[0041] For example, if the road marking layout data includes dashed line offset values ​​and dashed line scaling values, then the connection points of road dashed lines in connected road segment units can be aligned based on the predicted dashed line offset values ​​and scaling values ​​of each road segment unit. That is, the offset of the road dashed lines at the connection points of adjacent road segment units relative to the baseline is determined based on the dashed line offset values, thereby controlling the offset of the road dashed lines of adjacent road segment units within an acceptable range, ensuring that the road dashed lines of adjacent road segment units do not misalign at the connection points. Furthermore, the scaling ratio of the road dashed lines in each road segment unit relative to a preset dashed line reference size is determined based on the dashed line scaling values. For example, in a roundabout road, the size of the road dashed lines of a certain section of the middle ring road can be set as the reference size. According to perspective, the road dashed lines of road segment units closer to the inner ring road are reduced in size relative to the reference size, while the road dashed lines of road segment units closer to the outer ring road are enlarged in size relative to the reference size, thereby achieving a continuous and smooth road dashed line display effect that conforms to visual perspective.

[0042] The aforementioned road marking alignment method extracts features from the data of a two-dimensional road network map and obtains the marking layout data corresponding to each road segment unit based on a pre-trained road marking alignment model. Then, based on the obtained marking layout data, it aligns the road markings in each road segment unit of the three-dimensional road network map when converting the two-dimensional road network map into a three-dimensional road network map, thereby improving the display effect of road markings in the three-dimensional road network map.

[0043] In one embodiment, refer to Figure 2 As shown, Figure 2 A flowchart illustrating the steps of training a road marking alignment model in one embodiment is shown. The method for training the road marking alignment model may include the following steps:

[0044] S1: Obtain a sample two-dimensional road network map.

[0045] S2: Extract planar features from each sample road segment unit in the sample two-dimensional road network map.

[0046] S3: Obtain the actual road condition data of each sample road segment unit in the sample two-dimensional road network map.

[0047] S4: Train the road marking alignment model based on the planar features of each sample road segment unit and actual road condition data.

[0048] In this embodiment, any one or more historically generated two-dimensional road network maps can be used as samples for training the road marking alignment model. The sample two-dimensional road network map includes two-dimensional data of at least one sample road segment unit from at least one road. The actual traffic data of the sample two-dimensional road network map refers to the measured real data of each road segment unit in the road network corresponding to that sample two-dimensional road network map.

[0049] For example, the actual road condition data includes at least one of the following: the relative positional relationship between each sample road segment unit, the road type of each sample road segment unit, and the marking type of each sample road segment unit. The relative positional relationship may include whether they are adjacent, connected, or relative to the preceding, following, left, and right road segment units, etc.; the road type may include roundabout roads, loop roads, one-way streets, highways, etc.; and the marking type may include dashed road lines, solid road lines, road boundary lines, straight or turning indicator lines, etc.

[0050] Specifically, the terminal can select any one or more historically generated two-dimensional road network maps as training samples, and can retrieve the actual road condition data corresponding to the sample two-dimensional road network map by requesting third-party application services, etc. The planar features of each sample road segment unit are used as model input variables, and the actual road condition data corresponding to each sample road segment unit are used as model target variables. The road marking alignment model is trained through machine learning.

[0051] In one embodiment, reference Figure 2 As shown, the road marking alignment model is trained based on the planar features of each sample road segment unit and actual road data, including:

[0052] S41: Generate the current lane marking layout data for each sample road segment unit according to the preset value range.

[0053] S42: Score the current road marking layout data based on actual road condition data.

[0054] S43: Train the road marking alignment model based on the scoring results.

[0055] In this embodiment, users can customize the upper and lower limits of the value range of the road marking layout data according to business needs. Specifically, the terminal randomly generates the current road marking layout data within the value range of each road marking layout data. The terminal performs 3D modeling based on the current road marking layout data of each sample road segment unit, and judges whether the currently built 3D map conforms to the real road conditions based on the actual road data corresponding to each sample road segment unit, thereby scoring the current road marking layout data corresponding to each sample road segment unit. Furthermore, information such as the positional perspective relationship of each road segment unit can also be combined during the scoring process. Based on the scoring results, the terminal can further adjust the current road marking layout data to obtain a model that can predict road marking layout data that meets the requirements.

[0056] In one embodiment, training the road marking alignment model based on the scoring results may include the following steps:

[0057] S431: Associate each sample road segment unit whose score in the current road marking layout data is greater than a preset threshold with its current road marking layout data.

[0058] S432: Adjust the current road marking layout data of each sample road segment unit whose current road marking layout data score is less than the preset threshold, and use the adjusted data as the current road marking layout data, and proceed to the step of scoring each current road marking layout data according to the actual road condition data (S42).

[0059] In this embodiment, current road marking layout data with a score greater than a preset threshold indicates that it is close to or conforms to the layout of road markings in real road conditions, while current road marking layout data with a score less than the preset threshold indicates that it does not conform to the layout of road markings in real road conditions. Specifically, the terminal can filter sample road segment units with scores less than the preset threshold based on the scores of the current road marking layout data, and adjust and optimize the corresponding current road marking layout data of the selected sample road segment units by modifying their current road marking layout data. For sample road segment units with current road marking layout data scores greater than the preset threshold, the current road marking layout data of the sample road segment unit can be directly retained, and the association between the current road marking layout data and the planar features of the sample road segment unit can be completed.

[0060] For example, when there are multiple current road marking layout data, each type of current road marking layout data can be scored separately. If the score of one or more current road marking layout data is less than a preset threshold, only the current road marking layout data of that type for that road segment unit can be adjusted. For example, if the current road marking layout data includes two types: current road marking offset value and current road marking scaling ratio value, and the score of the current road marking offset value is greater than a preset threshold while the score of the current road marking scaling ratio value is less than a preset threshold, then only the current road marking scaling ratio value for that road segment unit can be adjusted, while the current road marking offset value for that road segment unit remains unchanged.

[0061] For example, when there are multiple current road marking layout data, a comprehensive score can be calculated based on the various current road marking layout data. If the comprehensive score of a road segment unit is less than a preset threshold, then the multiple current road marking layout data of that road segment unit will be adjusted as a whole.

[0062] Through this embodiment, sample road segment units that need to be adjusted can be dynamically selected based on the score of the current road marking layout data. Sample road segment units whose scores of the current road marking layout data are greater than a preset threshold do not need to be processed repeatedly, thereby reducing the amount of data processing. Therefore, targeted training and adjustment can be achieved on a per-road segment unit basis.

[0063] Furthermore, the current road marking layout data for each sample road segment unit with a score lower than a preset threshold is adjusted, and the adjusted new road marking layout data is then used as the current road marking layout data in the scoring process. By continuously iterating through scoring, filtering, and adjustment, the optimal road marking alignment model can be obtained when all road marking layout data for all sample road segment units meet the expected requirements.

[0064] In one embodiment, the method further includes: displaying a three-dimensional road network map on a display interface; and, in response to an adjustment operation on the marking layout data of a specified road segment unit in the three-dimensional road network map, returning the adjusted marking layout data to a road marking alignment model for model optimization.

[0065] In this embodiment, the terminal can display the converted 3D road network map through a display screen or other display device, and supports data modification and fine-tuning through user interaction. After responding to the user's adjustment operation on the layout data of the markings of the specified road segment unit, the adjusted data can be used as model optimization data to further feed back to the road marking alignment model, thereby optimizing the model and improving the prediction accuracy of the model.

[0066] It should be understood that, although Figures 1-2 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figures 1-2 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0067] In one embodiment, such as Figure 3 As shown, a road marking alignment device is provided, including: a two-dimensional map acquisition module 310, a planar feature extraction module 320, a layout data prediction module 330, and a road marking alignment module 340, wherein:

[0068] The 2D map acquisition module 310 is used to acquire a 2D road network map;

[0069] The planar feature extraction module 320 is used to extract planar features from each road segment unit in the two-dimensional road network map.

[0070] The layout data prediction module 330 is used to predict the road marking layout data of each road segment unit based on the extracted planar features of each road segment unit and the pre-trained road marking alignment model. The road marking alignment model is a model that characterizes the relationship between the planar features of each road segment unit and the road marking layout data.

[0071] The road marking alignment module 340 is used to align the road markings in each road segment unit according to the predicted marking layout data of each road segment unit when converting a two-dimensional road network map into a three-dimensional road network map.

[0072] In one embodiment, the system further includes a model training module 350, which is used to acquire a sample two-dimensional road network map; extract planar features from each sample road segment unit in the sample two-dimensional road network map; acquire actual road condition data of each sample road segment unit in the sample two-dimensional road network map; and train a road marking alignment model based on the planar features of each sample road segment unit and the actual road condition data.

[0073] In one embodiment, the model training module 350 generates current lane marking layout data for each sample road segment unit according to a preset value range; scores each current lane marking layout data according to actual road condition data; and trains the road marking alignment model based on the scoring results.

[0074] In one embodiment, the model training module 350 associates each sample road segment unit whose score in the current road marking layout data is greater than a preset threshold with its current road marking layout data; adjusts the current road marking layout data of each sample road segment unit whose score in the current road marking layout data is less than the preset threshold, and uses the adjusted data as the current road marking layout data, and proceeds to the step of scoring each current road marking layout data according to the actual road condition data.

[0075] In one embodiment, the road marking alignment module 340 is also used to display the three-dimensional road network map on the display interface; in response to the adjustment operation of the marking layout data of the specified road segment unit in the three-dimensional road network map, the adjusted marking layout data is returned to the road marking alignment model for model optimization.

[0076] Specific limitations regarding the road marking alignment device can be found in the limitations of the road marking alignment method described above, and will not be repeated here. Each module in the aforementioned road marking alignment device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0077] The road marking alignment method of this application will be further explained below with an application example. (Refer to...) Figure 4 As shown, Figure 4 A block diagram of a road marking alignment model for an application example is shown. The model training module 350 may include a training component 3502, an evaluation component 3504, and a loop adjustment component 3506.

[0078] More specifically, the following steps can be achieved based on the model training module 350:

[0079] Step 1: The training component 3502 acquires a sample 2D road network map. This sample 2D map includes multiple roads, each of which is further divided into at least one sample road segment unit. The training component 3502 generates a 3D road network map based on the planar features of each sample road segment unit in the sample 2D road network map. During the 3D road network map construction process, the training component 3502 randomly provides the lane markings for each sample road segment unit within a preset value range.

[0080] Step 2: The cyclic adjustment component 3506 inputs the lane marking layout data of each sample road segment unit output by the training component 3502 into the evaluation component 3504 for scoring.

[0081] Step 3: Evaluation component 3504 scores the marking layout data of each sample road segment unit in the 3D road network map based on the actual road data corresponding to the input 2D road network map (e.g., the relative positional relationship between each sample road segment unit, the road type of each sample road segment unit, the marking type of each sample road segment unit, etc.).

[0082] Step 4: The cyclic adjustment component 3506 feeds the scoring results back to the training component 3502. A higher score for the lane marking layout data indicates that the 3D visualization of the sample road segment unit is closer to the real situation. The training component 3502 retains all or part of the lane marking layout data.

[0083] Step 5: The training component 3502 adjusts the parameters of the initially generated 3D road network map based on the scoring results returned by the loop adjustment component 3506. During adjustment, the marking layout data of sample road segment units with scores greater than the threshold are kept unchanged, while the marking layout data of sample road segment units with scores below the threshold are optimized or adjusted.

[0084] Step 6: Input the adjusted pavement layout data of the sample road segment units back into the evaluation component 3504 for scoring. Repeat this process until the scores of all types of pavement layout data for all sample road segment units meet the preset requirements.

[0085] Step 7: Based on steps 1 to 6 above, obtain the trained road marking alignment model.

[0086] Step 8: The layout data prediction module 330 uses the trained road marking alignment model to predict the road marking layout data of each road segment unit in any one or more target two-dimensional road network maps to be converted, and through the road marking alignment module 340, when converting the target two-dimensional road network map into a three-dimensional road network map, it performs alignment processing on the road markings of each road segment unit in the three-dimensional road network map.

[0087] In one embodiment, a computer device is also provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a road network map processing method. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse.

[0088] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0089] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it performs the following steps: acquiring a two-dimensional road network map; extracting planar features from each road segment unit in the two-dimensional road network map; predicting the road marking layout data of each road segment unit based on the extracted planar features of each road segment unit and a pre-trained road marking alignment model, wherein the road marking alignment model is a model representing the correlation between the planar features of each road segment unit and the road marking layout data; and aligning the road markings in each road segment unit according to the predicted road marking layout data when converting the two-dimensional road network map into a three-dimensional road network map.

[0090] In one embodiment, when the processor executes the computer program, it also implements the step of training the road marking alignment model, specifically implementing the following steps: obtaining a sample two-dimensional road network map; extracting planar features from each sample road segment unit in the sample two-dimensional road network map; obtaining actual road condition data for each sample road segment unit in the sample two-dimensional road network map; and training the road marking alignment model based on the planar features of each sample road segment unit and the actual road condition data.

[0091] In one embodiment, when the processor executes a computer program to train the road marking alignment model based on the planar features of each sample road segment unit and actual road condition data, the following steps are specifically implemented: generating current marking layout data for each sample road segment unit according to a preset value range; scoring each current marking layout data according to actual road condition data; and training the road marking alignment model based on the scoring results.

[0092] In one embodiment, when the processor executes a computer program to train the road marking alignment model based on the scoring results, it specifically implements the following steps: associating each sample road segment unit whose score of the current marking layout data is greater than a preset threshold with its current marking layout data; adjusting the current marking layout data of each sample road segment unit whose score of the current marking layout data is less than the preset threshold, and using the adjusted data as the current marking layout data, and proceeding to the step of scoring each current marking layout data based on actual road condition data.

[0093] In one embodiment, when the processor executes the computer program, it further performs the following steps: displaying a three-dimensional road network map on a display interface; and, in response to an adjustment operation on the marking layout data of a specified road segment unit in the three-dimensional road network map, returning the adjusted marking layout data to the road marking alignment model for model optimization.

[0094] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon. When executed by a processor, the computer program performs the following steps: acquiring a two-dimensional road network map; extracting planar features from each road segment unit in the two-dimensional road network map; predicting the road marking layout data of each road segment unit based on the extracted planar features of each road segment unit and a pre-trained road marking alignment model, wherein the road marking alignment model is a model representing the correlation between the planar features of each road segment unit and the road marking layout data; and aligning the road markings in each road segment unit according to the predicted road marking layout data when converting the two-dimensional road network map into a three-dimensional road network map.

[0095] In one embodiment, when the computer program is executed by the processor, it also implements the step of training a road marking alignment model, specifically implementing the following steps: obtaining a sample two-dimensional road network map; extracting planar features from each sample road segment unit in the sample two-dimensional road network map; obtaining actual road condition data for each sample road segment unit in the sample two-dimensional road network map; and training the road marking alignment model based on the planar features of each sample road segment unit and the actual road condition data.

[0096] In one embodiment, when a computer program is executed by a processor to train a road marking alignment model based on the planar features of each sample road segment unit and actual road condition data, the following steps are specifically implemented: generating current marking layout data for each sample road segment unit according to a preset value range; scoring each current marking layout data according to actual road condition data; and training the road marking alignment model based on the scoring results.

[0097] In one embodiment, when a computer program is executed by a processor to train a road marking alignment model based on the scoring results, the following steps are specifically implemented: associating each sample road segment unit whose score in the current marking layout data is greater than a preset threshold with its current marking layout data; adjusting the current marking layout data of each sample road segment unit whose score in the current marking layout data is less than the preset threshold, and using the adjusted data as the current marking layout data, and proceeding to the step of scoring each current marking layout data based on actual road condition data.

[0098] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: displaying the three-dimensional road network map on the display interface; and, in response to an adjustment operation on the marking layout data of a specified road segment unit in the three-dimensional road network map, returning the adjusted marking layout data to the road marking alignment model for model optimization.

[0099] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0100] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0101] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A road marking alignment method, the method comprising: obtaining a two-dimensional road network map; extracting planar features of each road segment unit in the two-dimensional road network map; predicting road marking arrangement data of each road segment unit according to the extracted planar features of each road segment unit and a pre-trained road marking alignment model, the road marking alignment model being a model representing the correlation between the planar features of each road segment unit and the road marking arrangement data; aligning road markings in each road segment unit according to the predicted road marking arrangement data of each road segment unit when converting the two-dimensional road network map into a three-dimensional road network map, wherein the road marking arrangement data comprises a road marking offset value and a road marking scaling ratio value, the road segment unit is a road segment unit from a roundabout road, and the aligning road markings in each road segment unit according to the predicted road marking arrangement data of each road segment unit comprises: setting the size of road markings in a certain road segment unit in an inner ring road as a reference size, and according to perspective relationship information, road markings in road segment units closer to the inner ring road are processed to be smaller relative to the reference size, and road markings in road segment units closer to an outer ring road are processed to be larger relative to the reference size.

2. The method of claim 1, wherein, The road marking arrangement data further comprises at least one of a road marking offset range value and a road marking scaling ratio range value.

3. The method of claim 1, wherein, The training method of the road marking alignment model comprises: obtaining a sample two-dimensional road network map; extracting planar features of each sample road segment unit in the sample two-dimensional road network map; obtaining actual road condition data of each sample road segment unit in the sample two-dimensional road network map; training the road marking alignment model according to the planar features of each sample road segment unit and the actual road condition data.

4. The method of claim 3, wherein, The actual road condition data comprises at least one of a positional relative relationship between each sample road segment unit, a road type of each sample road segment unit, and a road marking type of each sample road segment unit.

5. The method of claim 3, wherein, The training of the road marking alignment model according to the planar features of each sample road segment unit and the actual road condition data comprises: generating current road marking arrangement data of each sample road segment unit according to a preset value range; scoring each current road marking arrangement data according to the actual road condition data; training the road marking alignment model according to the scoring results.

6. The method of claim 5, wherein, The training of the road marking alignment model according to the scoring results comprises: associating each sample road segment unit with its current road marking arrangement data when the score of the current road marking arrangement data of the sample road segment unit is greater than a preset threshold value; adjusting the current road marking arrangement data of each sample road segment unit when the score of the current road marking arrangement data of the sample road segment unit is less than the preset threshold value, and using the adjusted data as the current road marking arrangement data to enter the step of scoring each current road marking arrangement data according to the actual road condition data.

7. The method according to any one of claims 1 to 6, characterized in that, The method further comprises: displaying the three-dimensional road network map on a display interface. In response to an adjustment operation on the alignment data of a specified road segment unit in the three-dimensional road network map, the adjusted alignment data is returned to the road marking alignment model for model optimization.

8. A road marking alignment device, characterized by The device comprises: a two-dimensional map acquisition module configured to acquire a two-dimensional road network map; a plane feature extraction module configured to perform plane feature extraction on each road segment unit in the two-dimensional road network map; an arrangement data prediction module configured to predict alignment data of each road segment unit according to the extracted plane feature of each road segment unit and a pre-trained road marking alignment model, the road marking alignment model being a model representing the correlation between the plane feature and the alignment data of each road segment unit; a road marking alignment module configured to align road markings in each road segment unit according to the predicted alignment data of each road segment unit when converting the two-dimensional road network map into a three-dimensional road network map; wherein the alignment data comprises a marking offset value and a marking scaling ratio value, the road segment unit is a road segment unit from a roundabout road, and the road marking alignment module is specifically configured to: set the size of road markings in a certain road segment unit in an inner ring road as a reference size, and perform size reduction processing on road markings of road segment units closer to the inner ring road relative to the reference size and size amplification processing on road markings of road segment units closer to an outer ring road relative to the reference size according to perspective relationship information.

9. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the steps of the method of any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 7.

Images