Road network automatic updating method and device based on signaling data, equipment, storage medium and program product

Abstract

The invention discloses an automatic road network updating method and device based on signaling data, equipment, a storage medium and a program product, and relates to the technical field of road network updating, and the method comprises the steps: obtaining a user signaling track, mapping the user signaling track to a regular grid, and obtaining a preset feature synthesis image; performing road center line extraction on the preset feature synthesis graph by using a deep learning model to obtain a track road network; performing road screening based on the track road network to obtain a target newly-added road; the existing road is hooked through the target newly-added road and the track road network, the road network is automatically updated, signaling big data and a road network extraction model of a deep convolutional network are fused, roads with different flows are adaptively extracted, automatic hooking of the newly-added road and the existing road is achieved, and the traffic flow is improved. The link of realizing road hooking through manual operation is replaced, and the road network updating efficiency is greatly improved. And meanwhile, fine inspection of the vector problem of road network hooking is realized, and the stability of the road network quality is ensured.

Description

Technical Field

[0001] The present application relates to the technical field of road network updating, and in particular to a method, device, equipment, storage medium and program product for automatic road network updating based on signaling data. Background Art

[0002] Road network update: refers to updating the newly added roads in the real world to the road network topology data for use in map applications such as travel route planning, driving navigation, and pick-up and drop-off point recommendations. Road network topology data is usually expressed and stored in the form of geographic information system points, lines, and surfaces. With the rapid development of my country's social economy, the scale of urban roads has grown rapidly, and the road topology has also changed rapidly. How to update the newly added roads to the existing road network in a comprehensive, fast, and accurate manner has become an important but difficult task. The existing method of detecting road network changes through remote sensing images has problems such as high data acquisition costs and long update cycles, which makes it difficult to meet people's urgent needs for realistic road topology data. The existing road network update method based on trajectory big data requires manual reference to crowdsourcing images, remote sensing images, and other auxiliary materials to make judgments one by one. The process is cumbersome and the judgment quality is unstable.

[0003] Most existing methods use trajectory clustering or trajectory matching to identify and extract new roads. However, due to the positioning accuracy of the trajectory itself and signal drift, many pseudo roads are easily generated, which interfere with the extraction of new roads and lead to low accuracy of the new road network. Summary of the invention

[0004] The main purpose of this application is to provide a method, device, equipment, storage medium and program product for automatic updating of a road network based on signaling data, aiming to solve the technical problem of low accuracy of existing road network updates.

[0005] To achieve the above purpose, the present application proposes a method for automatically updating a road network based on signaling data, and the method for automatically updating a road network based on signaling data includes:

[0006] Acquire a user signaling trajectory, and map the user signaling trajectory onto a regular grid to obtain a preset feature composite image;

[0007] Use a deep learning model to extract the road centerline from the preset feature synthetic image to obtain a trajectory road network;

[0008] Perform road screening based on the trajectory road network to obtain target newly added roads;

[0009] The existing roads are connected through the target newly added roads and the track road network to complete the automatic update of the road network.

[0010] In one embodiment, the step of screening roads based on the trajectory road network to obtain target newly added roads includes:

[0011] Obtain existing road network data for existing roads;

[0012] Calculating the similarity between the trajectory road network and the existing road network data to obtain a similarity value;

[0013] The line segments in the trajectory road network whose similarity values ​​are greater than a preset similarity threshold are used as newly added line segments to obtain an initial newly added road;

[0014] Confidence calculation is performed on the initial newly added roads, and target newly added roads are obtained from the initial newly added roads.

[0015] In one embodiment, the step of calculating the similarity between the trajectory road network and the existing road network data to obtain a similarity value comprises:

[0016] Performing equidistant interpolation on line segments in the trajectory road network and line segments in the existing road network data to obtain trajectory road segment interpolation points, existing road interpolation points, and target distances between the trajectory road segment interpolation points and the existing road interpolation points;

[0017] Obtaining the average angle of the line segment after interpolation of the trajectory road segment interpolation point and the existing road interpolation point;

[0018] Get the width value and preset weight value of the existing road;

[0019] The similarity between the trajectory road network and the existing road network data is calculated by using the preset weight value, the average angle of the line segment, the target distance and the width value to obtain a similarity value.

[0020] In one embodiment, the step of calculating the confidence of the initial newly added roads and selecting target newly added roads from the initial newly added roads comprises:

[0021] Performing equidistant interpolation on the initially added roads on the regular grid to obtain a plurality of interpolation points;

[0022] A window with a preset side length is defined with the pixel of each interpolation point as the center to obtain multiple windows;

[0023] Count the total pixel flow in each window;

[0024] Obtaining the number of trajectories of the user signaling trajectory passing through the regular grid to obtain a pixel flow value;

[0025] Calculate the cross-sectional flow value of the initial newly added road by using the pixel flow sum and the pixel flow value;

[0026] Calculate the median of the cross-sectional flow value to obtain the newly added road flow value;

[0027] Get the historical newly added road flow value;

[0028] Calculating the flow value change by using the historical newly added road flow value and the newly added road flow value;

[0029] The initial newly added roads corresponding to the flow value change exceeding the preset change value are screened as target newly added roads.

[0030] In one embodiment, the step of linking the existing road with the target newly added road and the track road network to complete the automatic road network update includes:

[0031] Mapping the target newly added road, the removed roads in the trajectory road network, and the existing roads onto a preset map sheet to obtain target newly added road pixels, removed road pixels, and existing road pixels;

[0032] Convert the target newly added road pixels, the removed road pixels and the existing road pixels into vector lines to obtain target newly added lines, removed lines and reference lines;

[0033] Using a depth-first traversal strategy to traverse the target newly added lines, the removed lines, and the reference lines to obtain an initial road cluster;

[0034] Performing anomaly detection and pruning on the initial road clusters to obtain target road clusters;

[0035] The target newly added lines and removed lines in the target road cluster are attached to the existing roads to complete the automatic update of the road network.

[0036] In one embodiment, the step of performing anomaly detection and pruning on the initial road cluster to obtain the target road cluster includes:

[0037] Eliminate line segments with overlapping endpoints and overlapping lines in the initial road cluster to obtain a reference road cluster;

[0038] Performing equidistant interpolation on the target newly added lines, the removed lines and the reference lines in the reference road cluster to obtain interpolation points, and obtaining angle values ​​of the road sections corresponding to the interpolation points and the distances between the interpolation points;

[0039] Abnormal roads in the angle values ​​and the distances that do not conform to preset road rules are pruned to obtain target road clusters.

[0040] In addition, to achieve the above purpose, the present application also proposes a road network automatic updating device based on signaling data, and the road network automatic updating device based on signaling data includes:

[0041] An acquisition module, used to acquire a user signaling trajectory and map the user signaling trajectory onto a regular grid to obtain a preset feature composite image;

[0042] An extraction module, used to extract the road centerline from the preset feature composite image using a deep learning model to obtain a trajectory road network;

[0043] A screening module, used for screening roads based on the trajectory road network to obtain target newly added roads;

[0044] The attachment module is used to attach the existing roads through the target newly added roads and the track road network to complete the automatic update of the road network.

[0045] In addition, to achieve the above-mentioned purpose, the present application also proposes a road network automatic update device based on signaling data, the device comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the road network automatic update method based on signaling data as described above.

[0046] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the road network automatic update method based on signaling data as described above are implemented.

[0047] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the method for automatic road network update based on signaling data as described above.

[0048] One or more technical solutions proposed in this application include: obtaining user signaling trajectories, and mapping the user signaling trajectories onto a regular grid to obtain a preset feature composite image; using a deep learning model to extract the road centerline of the preset feature composite image to obtain a trajectory road network; screening roads based on the trajectory road network to obtain target new roads; connecting existing roads with the target new roads and the trajectory road network to complete automatic road network updates, integrating signaling big data and a deep convolutional network road network extraction model to adaptively extract roads with different traffic volumes, while suppressing multiple trajectory noises, and realizing automatic connection between new roads and existing roads, replacing the link of realizing road connection through manual operation, and greatly improving the efficiency of road network updates. At the same time, a detailed inspection of vector problems in road network connection is realized to ensure the stability of road network quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0050] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0051] Figure 1 A flowchart of the first embodiment of the method for automatic updating of a road network based on signaling data provided by the present application;

[0052] Figure 2 This is a schematic diagram of the user signaling trajectory modeling and road network inference process in an embodiment of the road network automatic update method based on signaling data of the present application;

[0053] Figure 3 A flowchart diagram of the second embodiment of the method for automatic updating of a road network based on signaling data provided by the present application;

[0054] Figure 4 This is a schematic diagram of the road matching and similarity calculation process in an embodiment of the method for automatic updating of a road network based on signaling data of the present application;

[0055] Figure 5 A flowchart of the third embodiment of the method for automatic updating of a road network based on signaling data provided by the present application;

[0056] Figure 6 This is a schematic diagram of the target new automatic attachment process in an embodiment of the method for automatic updating of a road network based on signaling data of the present application;

[0057] Figure 7This is a schematic diagram of the module structure of a device for automatically updating a road network based on signaling data according to an embodiment of the present application;

[0058] Figure 8 This is a schematic diagram of the device structure of the hardware operating environment involved in the method for automatic road network update based on signaling data in an embodiment of the present application.

[0059] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0060] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0061] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0062] The main solution of the embodiment of the present application is: obtain user signaling trajectory, and map the user signaling trajectory onto a regular grid to obtain a preset feature composite image; use a deep learning model to extract the road centerline of the preset feature composite image to obtain a trajectory road network; perform road screening based on the trajectory road network to obtain target new roads; connect the existing roads through the target new roads and the trajectory road network to complete automatic update of the road network.

[0063] Most existing technologies use trajectory clustering or trajectory matching to realize the identification and extraction of new roads. However, due to the positioning accuracy of the trajectory itself and signal drift, it is easy to generate many pseudo roads, which interfere with the extraction of new roads. The existing methods do not pay enough attention to suppressing trajectory noise, resulting in many pseudo roads. Moreover, the traffic volume of different roads varies greatly. The existing methods cannot take into account the road recognition tasks in various scenarios such as main roads and side roads, resulting in the inability to eliminate pseudo roads. And there is no evaluation of the quality of new roads. Only roads with clear spatial paths and stable traffic volume are new roads that are valuable to users. However, the existing methods do not perform confidence grading on new roads, eliminate suspected false roads, roads with chaotic paths, and roads with long-term low traffic volume, and it is very easy to update low-value roads to the existing road network, reducing the overall stability of the road network. There is no automatic update of new roads to the existing road network. The volume of new roads, existing roads and incremental roads in the country is huge. The existing methods need to rely on manual reference to crowdsourced images and remote sensing images to determine how to connect the new roads to the existing road network. The process is complicated and lengthy, and the cost is high. The update speed is slow and cannot keep up with the growth rate of the actual road network.

[0064] This application provides a solution. The method first maps the trajectory onto a regular grid and uses a CNN model to extract the centerline of the road. Then, a comprehensive similarity measurement algorithm is used to extract the part of the trajectory-generated road network with low similarity to the existing road network to obtain the newly added roads. Then, a historical comprehensive confidence algorithm is used to screen out high-confidence newly added roads with clear spatial positions, stable in time, and continuously maintained at a high level of traffic. Finally, a newly added road attachment algorithm based on vector-to-raster conversion and intelligent search is used to realize the automatic search and attachment of the connecting path between the newly added road network and the existing road network.

[0065] It should be noted that the execution subject of this embodiment may be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a road network automatic update device based on signaling data, etc. The following takes the road network automatic update device based on signaling data as an example to illustrate this embodiment and the following embodiments.

[0066] Based on this, the embodiment of the present application provides a method for automatically updating a road network based on signaling data, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the method for automatic updating of a road network based on signaling data of the present application.

[0067] In this embodiment, the method for automatically updating a road network based on signaling data includes steps S10 to S40:

[0068] Step S10: Acquire a user signaling trajectory, and map the user signaling trajectory onto a regular grid to obtain a preset feature composite image.

[0069] It should be noted that the user signaling trajectory is the user's driving trajectory within the road network that needs to be updated. The user's movement trajectory in the mobile communication network can be obtained by analyzing the user's mobile phone signaling data. The user signaling trajectory usually includes user ID, timestamp, cell identification code (CID), signaling event type, and base station latitude and longitude. By analyzing these data, the user's travel mode, stopover point, work and residence, travel method and other characteristics can be mined.

[0070] In a specific implementation, the user signaling trajectory can be mapped onto a regular grid to form a preset feature composite image, and the preset feature composite image is a pseudo-color image synthesized by multiple features.

[0071] In the specific implementation, the regular grid can be modeled in advance. Specifically, the whole country can be divided into squares of the same length and width according to the spatial position, and the squares do not overlap. Each square is further divided into 1024×1024 pixels. Taking a square with a length and width of 1.2 kilometers as an example, the resolution of each pixel is 1.2 meters×1.2 meters, thus obtaining a regular grid.

[0072] In a feasible implementation, the step of mapping the user signaling trajectory onto a regular grid in step S10 is specifically as follows: for each user signaling trajectory, a linear interpolation algorithm is used to calculate the grids it passes through. For the same trajectory, the speed value and direction value of the grids passed between two consecutive points are the same. Each grid counts the average speed, average direction and flow of all passing trajectories. In each individual block, the log function normalization method is used to normalize the average speed, average direction and flow to the range of 0-1. The specific calculation method is shown in Formula 1 below:

[0073] x'=log 10 (x) / log 10 (max) (Formula 1)

[0074] Where x is the average speed, average direction or flow of the pixel, max is the maximum value of the block feature, and x is the normalized feature value. Finally, the three features are synthesized into a pseudo-color image as the input of the deep learning model.

[0075] Step S20: Use a deep learning model to extract the road centerline of the preset feature composite image to obtain a trajectory road network.

[0076] In the specific implementation, the deep learning model is a CNN model. The CNN model is divided into three parts: encoder, central part and decoder. The encoder, central part and decoder are used to extract the road centerline from the preset feature synthetic image to obtain the trajectory road network. The specific steps are as follows: the encoder uses ResNet34 pre-trained on the ImageNet dataset, and the input is a preset feature synthetic image of 1024×1024. The feature map size of the encoder output is 32×32; the central part uses a hole convolution layer combined with series and parallel, and the expansion coefficients are 1, 2, 4, and 8. Therefore, 31×31 points on the first central feature map can be seen on the feature points on the last central layer, which multiplies the range of feature perception and integrates multi-scale features; the decoder uses a transposed convolution layer for upsampling to restore the resolution of the feature map from 32×32 to 1024×1024. The input of the model is a three-channel image with a resolution of 1024×1024. The value of the pixel point in the output image is the road probability value, ranging from 0 to 255. The closer the value is to 255, the higher the probability of being a road. Finally, the road grid is refined and shrunk, the single pixel of the central axis is extracted, and the grid along the central axis is vectorized to obtain a vector road network, namely, a trajectory road network.

[0077] like Figure 2 As shown, Figure 2 This is a schematic diagram of the user signaling trajectory modeling and road network inference process. By mapping the user signaling trajectory onto a regular grid that has been spatially segmented and grid-managed, the three-channel features of flow, speed, and direction are output to form a three-feature synthetic pseudo-color image, which is then input into a deep convolutional network model to output a probability map. The probability map represents the value of the pixel point in the image as a road probability value, ranging from 0 to 255. The closer the value is to 255, the higher the probability of being a road. Finally, the road grid is refined and shrunk, the single pixel of the central axis is extracted, and the grid along the central axis is vectorized to obtain a vector road network.

[0078] Step S30: Screening roads based on the trajectory road network to obtain target newly added roads.

[0079] In the specific implementation, some of the road networks generated by the trajectory are consistent with the existing roads, while some are inconsistent with the existing roads, that is, roads with low similarity to the existing roads. The trajectory road network can be screened through the existing roads to obtain the target new roads. The target new roads are roads with low similarity to the existing roads.

[0080] In specific implementations, similarity algorithms, clustering algorithms, etc. can be used to screen the trajectory road network to obtain target new roads.

[0081] Step S40: the existing roads are connected through the target newly added roads and the track road network to complete the automatic update of the road network.

[0082] It is understandable that after obtaining the target new roads, the existing roads can be connected according to the target new roads and the remaining roads in the trajectory road network except the target new roads, so as to connect the new roads with the existing roads, replacing the link of road connection through manual operation, greatly improving the efficiency of road network update. At the same time, the vector problem of road network connection is finely checked to ensure the stability of road network quality.

[0083] It should be noted that, by connecting the target newly added roads and the trajectory road network to the existing roads, the road network matching can be performed first, and then the topological connection can be performed, and finally the road network can be updated.

[0084] The road network matching can be geometric matching, which finds the nearest existing road by geometrically matching the target newly added road with the roads in the existing road network. Specifically, the distance between the start and end point of the newly added road and the start and end points of all roads in the existing road network can be calculated or the nearest existing road can be found by using a nearest neighbor search algorithm (such as KD Tree).

[0085] Topological connection may include endpoint connection and middle point connection. Endpoint connection is to connect the starting point and end point of the newly added road with the endpoint of the nearest road in the existing road network. Specifically, endpoint matching can be performed first, and then the specific connection method can be determined. For example, if the distance between the starting point or end point of the newly added road and the endpoint of the existing road is less than a certain threshold (such as 10 meters), it is considered that the two points can be connected. Straight line connection or curve connection can be used, depending on the actual terrain and road layout. Middle point connection is: if the middle part of the newly added road has an intersection with the existing road, these intersections are connected. Intersection detection can be performed first, and then the specific connection method can be determined. Intersection detection can use the line segment intersection algorithm to detect whether the newly added road and the existing road have an intersection. The connection method can be to use the intersection as a new node to connect the newly added road and the existing road.

[0086] Suppose there is a road network update task in an urban area, and a new road (target new road) needs to be connected to the existing road network.

[0087] Sample data: Existing road network: Path A: start point (0,0), end point (10,0); Path B: start point (10,0), end point (10,10); Path C: start point (10,10), end point (0,10). Target new road: New road D: start point (5,5), end point (15,5).

[0088] Data preparation: The existing road network includes paths A, B, and C. The target new road is path D.

[0089] Road network matching: Distance calculation: Calculate the distance between the starting point (5,5) of the newly added road D and the starting and ending points of all roads in the existing road network. Calculate the distance between the ending point (15,5) of the newly added road D and the starting and ending points of all roads in the existing road network. Nearest neighbor search: Use KD Tree to find the nearest existing road. For example, the nearest existing road to the starting point (5,5) of the newly added road D is the ending point (10,0) of path A. The nearest existing road to the ending point (15,5) of the newly added road D is the starting point (10,0) of path B.

[0090] Topological connection: Endpoint connection: The distance between the starting point (5,5) of the newly added road D and the end point (10,0) of path A is 5.83 meters, which is less than the threshold of 10 meters and can be connected. The distance between the end point (15,5) of the newly added road D and the starting point (10,0) of path B is 7.07 meters, which is less than the threshold of 10 meters and can be connected. Connection method: Use straight line connection to connect the starting point (5,5) of the newly added road D and the end point (10,0) of path A. Use straight line connection to connect the end point (15,5) of the newly added road D and the starting point (10,0) of path B.

[0091] Road network update: Update road network data: Add new nodes and edges to the graph data structure. Add nodes (5,5) and (15,5). Add edges (5,5)->(10,0) and (15,5)->(10,0).

[0092] Verification and optimization: Topology verification: Check whether the connection of the newly added road D destroys the topological structure of the road network and ensure that no isolated nodes or loops are formed. Path optimization: Optimize the path of the newly added road D to ensure the rationality and connectivity of the path.

[0093] The updated road network is as follows: Path A: starting point (0,0), end point (10,0); Path B: starting point (10,0), end point (10,10); Path C: starting point (10,10), end point (0,10); Path D: starting point (5,5), end point (15,5); connection: (5,5)->(10,0), (15,5)->(10,0). Through the above steps, the target newly added road D is successfully connected to the existing road network, and the automatic update of the road network is completed.

[0094] In specific implementation, the automatic road network update of this embodiment can be applied to different types of road updates, such as urban road updates, rural road updates, etc. If applied to urban road updates, the user signaling trajectory can be obtained by collecting the collected signaling data of urban users. These trajectories are mapped to a regular grid to form a preset feature composite image. The CNN model is used to extract the road centerline of the preset feature composite image to obtain a trajectory road network. The target new roads are screened out through a comprehensive similarity measurement algorithm and a historical comprehensive confidence algorithm. Finally, the target new roads are connected to the existing road network through the new road connection algorithm of intelligent search to complete the automatic update of the road network.

[0095] In a rural area, due to the low traffic volume, it is difficult for existing methods to effectively identify new roads. By collecting user signaling data in the area, user signaling trajectories are obtained. These trajectories are mapped to a regular grid to form a preset feature composite image. The CNN model is used to extract the road centerline of the preset feature composite image to obtain a trajectory road network. The target new roads are screened out through the comprehensive similarity measurement algorithm and the historical comprehensive confidence algorithm. Finally, the target new roads are connected to the existing road network through the new road connection algorithm of intelligent search to complete the automatic update of the road network.

[0096] This embodiment provides a method for automatically updating a road network based on signaling data, which obtains user signaling trajectories and maps the user signaling trajectories onto a regular grid to obtain a preset feature composite image; uses a deep learning model to extract the road centerline of the preset feature composite image to obtain a trajectory road network; screens roads based on the trajectory road network to obtain target newly added roads; connects existing roads with the target newly added roads and the trajectory road network to complete automatic road network update, integrates signaling big data and a deep convolutional network road network extraction model, adaptively extracts roads with different traffic volumes, and suppresses multiple trajectory noises, thereby achieving automatic connection between newly added roads and existing roads, replacing the link of achieving road connection through manual operation, and greatly improving the efficiency of road network update. At the same time, a detailed inspection of vector problems in road network connection is achieved to ensure the stability of road network quality.

[0097] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 3 , step S30 includes steps S301 to S304:

[0098] Step S301: Acquire existing road network data of existing roads.

[0099] It should be noted that the existing road network data of the existing road on the regular road network can be obtained, and specifically, the existing road network data corresponding to the existing road on the regular road network can be obtained according to the location, longitude and latitude and other information of the existing road.

[0100] Step S302: Calculate the similarity between the trajectory road network and the existing road network data to obtain a similarity value.

[0101] In a specific implementation, a comprehensive similarity algorithm may be used to calculate the similarity between the trajectory road network and the existing road network data, thereby obtaining a calculated similarity value. Other similarity calculation algorithms or clustering algorithms may also be used to calculate the similarity between the trajectory road network and the existing road network data, which is not limited in this embodiment.

[0102] For example, by extracting features from the trajectory road network and the existing road network, the geometric features, topological features, and traffic features of the roads can be obtained, and the similarity value can be obtained by calculating the geometric similarity, topological similarity, and traffic similarity between the two roads. The similarity value between the trajectory road network and the existing road network can also be calculated by equidistant interpolation to improve the accuracy and robustness of the similarity calculation.

[0103] In a feasible implementation, step S302 may include steps A11 to A14:

[0104] Step A11: performing equidistant interpolation on line segments in the trajectory road network and line segments in the existing road network data to obtain trajectory segment interpolation points, existing road interpolation points, and target distances between the trajectory segment interpolation points and the existing road interpolation points.

[0105] It should be noted that the line segments in the trajectory road network and the line segments in the existing road network data can be interpolated at equal distances, so as to obtain the trajectory segment interpolation points, the existing segment interpolation points and the shortest distances between the trajectory segment interpolation points and the existing segment interpolation points, and the target distance is the shortest distance.

[0106] like Figure 4 As shown, Figure 4 The schematic diagram of the road matching and similarity calculation process is shown in Figure 1. For example, a point is inserted every 2 meters. a , find the nearest existing road interpolation point P point by point b , the shortest distance is set to d, and the trajectory road interpolation point is P 11 , P 12 , P 13 , P 14. ..P 1n The existing road interpolation point is P 21 , P 22 , P 23 , P 24. ..P2n .

[0107] Step A12: Obtain the average angle of the line segment after interpolation of the trajectory segment interpolation points and the existing road interpolation points.

[0108] In a specific implementation, the interpolation point P can be calculated a With P b The average angle g of the two interpolated line segments.

[0109] Step A13: Obtain the width value and preset weight value of the existing road.

[0110] In a specific implementation, the width value of an existing road can be obtained by querying the road setting specifications, and the preset weight values ​​A and B can be set in advance and adjusted according to actual conditions.

[0111] Step A14: Calculate the similarity between the trajectory road network and the existing road network data by using the preset weight value, the average angle of the line segment, the target distance and the width value to obtain a similarity value.

[0112] It should be understood that the similarity between the trajectory network and the existing network data can be calculated by presetting the weight values ​​A and B, the average angle g of the line segment, the target distance d, and the width value D, as shown in the following formula 2:

[0113]

[0114] In the above formula 2, d is the interpolation point P a With P b The distance between the interpolation points P and D is the actual width of the roads of different grades. When g is 0 and d is 0, W is 0, indicating that the interpolation point P a With P b The corresponding road sections match perfectly.

[0115] Step S303: taking the line segments in the trajectory road network whose similarity values ​​are greater than a preset similarity threshold as newly added line segments to obtain an initial newly added road.

[0116] In a specific implementation, the preset similarity threshold may be set to 0.8, 0.9, etc., and this embodiment does not limit this. When the similarity value W exceeds the preset similarity threshold, it indicates that the interpolation point P a With P b The corresponding road segment does not match. For each road segment in the trajectory road network, calculate the similarity W of the existing road after interpolation, and extract the line segment with a W value greater than the threshold, which is the initial newly added road.

[0117] Step S304: Calculate the confidence of the initial newly added roads, and select target newly added roads from the initial newly added roads.

[0118] It is understandable that due to the positioning accuracy of the trajectory itself and signal drift, there are many pseudo roads in the newly added roads, which have no corresponding roads in the real world, so further filtering is required. The confidence of the initial newly added roads can be calculated, so as to screen the initial newly added roads and use the initial information roads with high confidence as the target newly added roads. The confidence calculation specifically adopts the historical comprehensive confidence algorithm to screen high-confidence roads with (1) clear spatial location and stable in time, and (2) high traffic level.

[0119] In a specific implementation, the confidence calculation can calculate the spatial confidence, temporal confidence, flow confidence, user behavior confidence, etc. of the road. When the initial newly added roads are screened based on the confidence, one or more of the spatial confidence, temporal confidence, flow confidence, and user behavior confidence can be considered. The comprehensive confidence can also be calculated based on the spatial confidence, temporal confidence, flow confidence, and user behavior confidence. The initial newly added roads can be screened by the comprehensive confidence to obtain the target new roads.

[0120] In a feasible implementation, step S304 may include steps B11 to B19:

[0121] Step B11: performing equidistant interpolation on the initially added roads on the regular grid to obtain a plurality of interpolation points.

[0122] It should be noted that the calculation of flow values ​​uses the pixel-scale flow grid map obtained in the trajectory modeling stage as a reference. The initial newly added roads are first interpolated at equal distances on the regular grid, and each interpolation point is mapped to a pixel of the flow grid map according to the longitude and latitude, thereby obtaining multiple interpolation points.

[0123] Step B12: A window with a preset side length is defined with the pixel of each interpolation point as the center to obtain multiple windows.

[0124] In a specific implementation, the pixel of each interpolation point can be taken as the center, and a preset side length of n, which generally corresponds to the average width of the regional road network, is used to delineate a square window with a side length of n, thereby obtaining multiple square windows.

[0125] Step B13: Count the total pixel flow in each window.

[0126] It should be noted that for each window, the total flow rate S of all pixels in the window is counted.

[0127] Step B14: Obtain the number of trajectories of the user signaling trajectory passing through the regular grid to obtain a pixel flow value.

[0128] In a specific implementation, the pixel flow value is the number of trajectories passing through the regular grid counted in the trajectory modeling stage, which can be obtained by counting on the regular grid.

[0129] Step B15: Calculate the cross-sectional flow value of the initial newly added road by using the pixel flow sum and the pixel flow value.

[0130] It should be understood that the cross-sectional flow value of the initial newly added road can be calculated by the pixel flow sum and the pixel flow value, and the cross-sectional flow value=pixel flow sum S / pixel flow value m.

[0131] Step B16: Calculate the median of the cross-sectional flow value to obtain the newly added road flow value.

[0132] In a specific implementation, after obtaining the cross-sectional flow value, the median of the cross-sectional flow value can be calculated, that is, the flow value of the newly added road can be obtained. The reason for using the median instead of the average value is that the average value is easily affected by the high flow at the intersections at both ends of the newly added road.

[0133] Step B17: Obtain the historical newly added road traffic value.

[0134] In specific implementations, the historical newly added road flow value is the newly added road flow value calculated in the previous period.

[0135] Step B18: Calculate the flow value change by using the historical newly added road flow value and the newly added road flow value.

[0136] It should be noted that the historical newly added road flow value and the newly added road flow value can be compared to calculate the difference between the flow values ​​and obtain the flow value change.

[0137] Step B19: Filter the initial newly added roads corresponding to the traffic value change exceeding the preset change value as target newly added roads.

[0138] The preset change value can be set in advance, such as 0.1, 0.2, etc. If the flow value change exceeds the preset change value, it means that the flow of the newly added road is maintained at a high level. The flow threshold here is found in actual production, and it is best to have at least 2 trajectories passing through the newly added road cross-section per week.

[0139] Among the newly added roads with low confidence in the current period, there are some newly built and open roads, which have low confidence because there is no information about this road in the historical data. In the subsequent calculations of each period, the confidence value of this road will be continuously updated. If a road becomes a high-confidence road in a certain period, it will also be extracted and entered into the next process.

[0140] In the specific implementation, in the process of screening target new roads, important new roads may be missed. Therefore, in addition to using signaling data, it is also possible to combine remote sensing images, satellite maps and other multi-source data to improve the accuracy and comprehensiveness of new road detection through cross-validation of multi-source data.

[0141] When processing roads, not only spatial features are considered, but also temporal features. For example, some roads may have higher traffic during certain time periods (such as morning and evening rush hours) and lower traffic at other times. By analyzing these spatiotemporal features, newly added roads can be identified more accurately.

[0142] At the same time, if a road maintains stable traffic for a long time, it will be targeted for addition even if the traffic is below the set threshold.

[0143] In the specific implementation, a user feedback mechanism can also be established. After the target new roads are obtained, the target new roads can be pushed to the users, so that the users can check whether there are any missed new roads, so as to iterate and update the screening rules.

[0144] This embodiment obtains the existing road network data of the existing roads; calculates the similarity between the trajectory road network and the existing road network data to obtain a similarity value; uses the line segments in the trajectory road network with similarity values ​​greater than a preset similarity threshold as newly added line segments to obtain the initial newly added roads; performs confidence calculation on the initial newly added roads, and selects the target newly added roads from the initial newly added roads. By calculating the similarity between the trajectory road network and the existing road network data, it is possible to identify the parts that may be omitted or not updated in the existing road network data, thereby improving the accuracy and completeness of the road data, and can quickly screen the newly added roads in the trajectory road network, and select high-confidence newly added roads with clear spatial positions, stable in time, and continuously maintained at a high level of traffic. By performing confidence calculation on the initial newly added roads, newly added roads with high data quality and strong reliability are selected to ensure the reliability and practicality of the road data, and confidence classification is performed on the newly added roads to eliminate false roads, roads with chaotic paths, and roads with low traffic, thereby improving the quality of the newly added roads from the source.

[0145] Based on the first embodiment of the present application, in the third embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 5 , step S40 includes steps S401 to S405:

[0146] Step S401: Mapping the target newly added road, the removed roads in the trajectory road network and the existing roads onto a preset map sheet to obtain target newly added road pixels, removed road pixels and existing road pixels.

[0147] It should be noted that, in addition to the target newly added roads, there are other eliminated roads that do not meet the requirements of newly added roads in the trajectory road network.

[0148] The difficulties of road connection include: (1) the connection path between the new roads and the existing road network is unclear, so how to find the best connection path? (2) how to deal with the complex topological relationship between the new roads and the existing road network, and automatically cluster the connected new roads. To address these problems, this paper proposes a new road connection algorithm based on vector-to-raster conversion and intelligent search.

[0149] The preset map is to splice adjacent 10×10 blocks into a larger map, thereby improving the reference information and success rate of the newly added road attachment in the block and reducing the parallel calculation amount of the attachment.

[0150] In the specific implementation, a blank canvas can be generated first, and the target newly added road is the high-confidence road, and the proposed road is the low-confidence road. By mapping the high-confidence road, low-confidence road and existing road to the blank canvas, high-confidence road pixels, low-confidence road pixels and existing road pixels are obtained. The existing road pixels directly cover the high, medium and low-confidence road pixels to achieve the erasure of the current road pixels.

[0151] Step S402: converting the target newly added road pixels, the removed road pixels and the existing road pixels into vector lines to obtain target newly added lines, removed lines and reference lines.

[0152] In a specific implementation, low-confidence road pixels, high-confidence road pixels, and existing road pixels can be converted into vector lines to obtain target added lines, removed lines, and reference lines. The reference lines are vector lines converted from existing road pixels.

[0153] Step S403: using a depth-first traversal strategy to traverse the target newly added lines, the removed lines and the reference lines to obtain initial road clusters.

[0154] In order to cluster the connected new roads, a recursive depth-first traversal algorithm is used to traverse the target new lines, removed lines, and reference lines to form the initial road clusters. Specifically, starting from the two end points of each high-confidence road, the medium and low confidence roads adjacent to the high-confidence road are searched, and then the adjacent medium and low confidence roads visited are used as the initial nodes to recursively search the nearby medium and low confidence roads until the existing roads are encountered. Finally, all the searched high, medium and low confidence roads are combined into the initial road clusters, and multiple sub-segments in the same road segment in the cluster are merged into a complete line.

[0155] Step S404: performing anomaly detection and pruning on the initial road clusters to obtain target road clusters.

[0156] In a specific implementation, after obtaining the initial road clusters, the existing roads can be connected according to the initial road clusters, and after the connection is completed, the internal topological relationship in the initial road clusters is detected and pruned for anomalies, so as to remove overlapping breakpoints and roads that do not conform to the rules and obtain the target road clusters.

[0157] Anomaly detection can identify and remove noise and false roads, improve the quality and accuracy of new roads, and specifically detect whether the position, angle, etc. between the target new road and the existing roads are abnormal.

[0158] In a feasible implementation, step S404 may include steps C11 to C13:

[0159] Step C11: removing line segments with overlapping endpoints and overlapping lines in the initial road clusters to obtain reference road clusters.

[0160] It should be noted that pruning includes detecting the overlap of line segment endpoints (the distance between endpoints is less than a threshold value such as 1 meter) and line overlap (the angle is less than a threshold value such as 15 degrees, and the distance is less than a threshold value such as 3 meters), and removing the overlapping breakpoints or lines to obtain reference road clusters. The reference road clusters are the road clusters after pruning.

[0161] Step C12: performing equidistant interpolation on the target newly added lines, removed lines and reference lines in the reference road cluster to obtain interpolation points, and obtaining angle values ​​of road sections corresponding to the interpolation points and distances between the interpolation points.

[0162] In the specific implementation, anomaly detection includes detecting the angle less than a threshold (such as 90 degrees) and detecting whether the target new road is parallel to the existing road (the angle is greater than a certain threshold, such as 150 degrees) and the distance is less than a threshold (such as 30 meters). In anomaly detection, the angle and distance are calculated as above Figure 4 As shown, first, the target line segment (high confidence line segment and low confidence line segment) and the reference line segment (existing line segment) are interpolated at equal distances (such as 2 meters). a , find the reference line segment interpolation point P closest to it point by point b , the shortest distance is set to d; interpolation point P a With P b The angle value of the corresponding road section is interpolated using point P a With P bThe average angle g of the two interpolated line segments is used as an approximation. The average value of the closest distance d obtained by all interpolation points of the target line segment is calculated, which is the distance value between the target line segment and the reference line segment; the average value of the angle g obtained by all interpolation points of the target line segment is calculated, which is the angle value between the target line segment and the reference line segment.

[0163] Step C13: Pruning abnormal roads in the angle value and the distance that do not conform to the preset road rules to obtain target road clusters.

[0164] In the specific implementation, the preset road rules define the angle value and distance between normal roads. The abnormal roads whose angle value and distance do not meet the preset road rules can be eliminated, and then the road clusters whose endpoints are attached to the existing road network are selected from the remaining sub-clusters as the final online road clusters. For deleted roads, if a new road cluster contains the road section in a future period, and all endpoints are attached to the existing road network, and no abnormalities are detected, they will also be pushed online.

[0165] Step S405: attaching the target newly added lines and removed lines in the target road cluster to the existing roads to complete the automatic update of the road network.

[0166] In the specific implementation, the target newly added lines and removed lines can be attached to the existing roads according to the topological relationship in the target road cluster, so as to complete the automatic update of the road network. Figure 6 As shown, Figure 6 Schematic diagram of the target new automatic attachment process, a is the relationship between existing roads, intelligence (target new roads) and CNN lines (removed roads), b is drawing the existing roads, target new roads and removed roads into existing road pixels, high-confidence road pixels and low-confidence road pixels, c is performing anomaly detection and pruning on the lines to form target road clusters, and d is extending each line in the target road cluster to attach it to the existing road.

[0167] This embodiment maps the target newly added roads, the removed roads in the trajectory road network, and the existing roads to a preset map sheet to obtain target newly added road pixels, removed road pixels, and existing road pixels; converts the target newly added road pixels, the removed road pixels, and the existing road pixels into vector lines to obtain target newly added lines, removed lines, and reference lines; uses a depth-first traversal strategy to traverse the target newly added lines, the removed lines, and the reference lines to obtain an initial road cluster; performs anomaly detection and pruning on the initial road cluster to obtain a target road cluster; and attaches the target newly added lines and removed lines in the target road cluster to the existing roads to complete the automatic update of the road network. Updating the road network data by an automated method can quickly reflect the actual changes in the roads, improve the accuracy and timeliness of the road network data, effectively reduce the manual attachment operation links in the road network production, and greatly improve the efficiency of road network updates. In addition, a vector problem fine inspection algorithm is used in the automatic attachment of the road network to ensure the stability of the road network quality.

[0168] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the automatic road network update method based on signaling data of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0169] This application also provides a road network automatic update device based on signaling data, please refer to Figure 7 , the automatic road network updating device based on signaling data comprises:

[0170] The acquisition module 10 is used to acquire the user signaling trajectory and map the user signaling trajectory onto a regular grid to obtain a preset feature composite image.

[0171] The extraction module 20 is used to use a deep learning model to extract the road centerline of the preset feature composite image to obtain a trajectory road network.

[0172] The screening module 30 is used to screen roads based on the trajectory road network to obtain target newly added roads.

[0173] The attachment module 40 is used to attach the existing roads through the target newly added roads and the track road network to complete the automatic update of the road network.

[0174] The automatic road network update device based on signaling data provided by the present application adopts the automatic road network update method based on signaling data in the above embodiment, which can solve the technical problem of low accuracy of existing road network updates. Compared with the prior art, the beneficial effects of the automatic road network update device based on signaling data provided by the present application are the same as the beneficial effects of the automatic road network update method based on signaling data provided by the above embodiment, and other technical features in the automatic road network update device based on signaling data are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0175] In one embodiment, the screening module 30 is also used to obtain existing road network data of existing roads; calculate the similarity between the trajectory road network and the existing road network data to obtain a similarity value; use the line segments in the trajectory road network whose similarity values ​​are greater than a preset similarity threshold as newly added line segments to obtain initial newly added roads; perform confidence calculation on the initial newly added roads, and screen the target newly added roads from the initial newly added roads.

[0176] In one embodiment, the screening module 30 is further used to perform equidistant interpolation on the line segments in the trajectory road network and the line segments in the existing road network data to obtain trajectory segment interpolation points, existing road interpolation points and target distances between the trajectory segment interpolation points and the existing road interpolation points; obtain the average angle of the line segments after interpolation of the trajectory segment interpolation points and the existing road interpolation points; obtain the width value and preset weight value of the existing road; calculate the similarity between the trajectory road network and the existing road network data through the preset weight value, the average angle of the line segments, the target distance and the width value to obtain a similarity value.

[0177] In one embodiment, the screening module 30 is further used to perform equidistant interpolation on the initial newly added roads on the regular grid to obtain multiple interpolation points; define a window of preset side length with the pixel of each interpolation point as the center to obtain multiple windows; count the sum of pixel flows in each window; obtain the number of trajectories of the user signaling trajectory passing through the regular grid to obtain a pixel flow value; calculate the cross-sectional flow value of the initial newly added roads by the sum of pixel flows and the pixel flow value; calculate the median of the cross-sectional flow value to obtain the flow value of the newly added road; obtain the historical flow value of the newly added road; calculate the flow value change by the historical flow value of the newly added road and the flow value of the newly added road; and screen the initial newly added roads corresponding to the flow value change exceeding the preset change value as target newly added roads.

[0178] In one embodiment, the attachment module 40 is further used to map the target newly added roads, the removed roads in the trajectory road network, and the existing roads onto a preset map sheet to obtain target newly added road pixels, removed road pixels, and existing road pixels; convert the target newly added road pixels, the removed road pixels, and the existing road pixels into vector lines to obtain target newly added lines, removed lines, and reference lines; use a depth-first traversal strategy to traverse the target newly added lines, the removed lines, and the reference lines to obtain initial road clusters; perform anomaly detection and pruning on the initial road clusters to obtain target road clusters; attach the target newly added lines and removed lines in the target road clusters to the existing roads to complete automatic road network update.

[0179] In one embodiment, the attachment module 40 is further used to remove line segments with overlapping endpoints and overlapping lines in the initial road cluster to obtain a reference road cluster; perform equidistant interpolation on the target newly added lines, the removed lines, and the reference lines in the reference road cluster to obtain interpolation points, and obtain the angle value of the road segment corresponding to the interpolation point and the distance between the interpolation points; prune abnormal roads that do not comply with preset road rules in the angle value and the distance to obtain a target road cluster.

[0180] The present application provides a road network automatic update device based on signaling data, and the road network automatic update device based on signaling data includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the road network automatic update method based on signaling data in the above-mentioned embodiment one.

[0181] Reference below Figure 8 , which shows a schematic diagram of the structure of a road network automatic update device based on signaling data suitable for implementing the embodiment of the present application. The road network automatic update device based on signaling data in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 8 The road network automatic update device based on signaling data shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0182] like Figure 8 As shown, the automatic road network update device based on signaling data may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. Various programs and data required for the operation of the automatic road network update device based on signaling data are also stored in RAM1004. The processing device 1001, ROM1002, and RAM1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems may be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 1003 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 1009. The communication device 1009 may allow the automatic update device for a road network based on signaling data to communicate wirelessly or wired with other devices to exchange data. Although the figure shows an automatic update device for a road network based on signaling data with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have alternatively.

[0183] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0184] The automatic road network update device based on signaling data provided by the present application adopts the automatic road network update method based on signaling data in the above embodiment, which can solve the technical problem of low accuracy of existing road network updates. Compared with the prior art, the beneficial effects of the automatic road network update device based on signaling data provided by the present application are the same as the beneficial effects of the automatic road network update method based on signaling data provided by the above embodiment, and other technical features in the automatic road network update device based on signaling data are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0185] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0186] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0187] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the automatic road network update method based on signaling data in the above-mentioned embodiment.

[0188] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0189] The computer-readable storage medium may be included in the automatic road network update device based on signaling data; or may exist independently without being assembled into the automatic road network update device based on signaling data.

[0190] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the road network automatic update device based on signaling data, the road network automatic update device based on signaling data: obtains the user signaling trajectory, and maps the user signaling trajectory onto a regular grid to obtain a preset feature composite image; uses a deep learning model to extract the road centerline of the preset feature composite image to obtain a trajectory road network; screens roads based on the trajectory road network to obtain target new roads; and connects existing roads through the target new roads and the trajectory road network to complete the automatic update of the road network.

[0191] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0192] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0193] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0194] The readable storage medium provided by the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned method for automatically updating a road network based on signaling data, and can solve the technical problem of low accuracy of existing road network updates. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as the beneficial effects of the method for automatically updating a road network based on signaling data provided by the above-mentioned embodiment, and will not be described in detail here.

[0195] The present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for automatic road network update based on signaling data.

[0196] The computer program product provided by the present application can solve the technical problem of low accuracy of existing road network updates. Compared with the prior art, the beneficial effects of the computer program product provided by the present application are the same as the beneficial effects of the road network automatic update method based on signaling data provided by the above embodiment, which will not be repeated here.

[0197] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for automatically updating a road network based on signaling data, characterized in that: The method for automatically updating a road network based on signaling data comprises: Acquire a user signaling trajectory, and map the user signaling trajectory onto a regular grid to obtain a preset feature composite image; Use a deep learning model to extract the road centerline from the preset feature synthetic image to obtain a trajectory road network; Perform road screening based on the trajectory road network to obtain target newly added roads; The existing roads are connected through the target newly added roads and the track road network to complete the automatic update of the road network.

2. The method according to claim 1, characterized in that The step of screening roads based on the trajectory road network to obtain target newly added roads includes: Obtain existing road network data for existing roads; Calculating the similarity between the trajectory road network and the existing road network data to obtain a similarity value; The line segments in the trajectory road network whose similarity values ​​are greater than a preset similarity threshold are used as newly added line segments to obtain an initial newly added road; Confidence calculation is performed on the initial newly added roads, and target newly added roads are obtained from the initial newly added roads.

3. The method according to claim 2, characterized in that The step of calculating the similarity between the trajectory road network and the existing road network data to obtain a similarity value comprises: Performing equidistant interpolation on line segments in the trajectory road network and line segments in the existing road network data to obtain trajectory road segment interpolation points, existing road interpolation points, and target distances between the trajectory road segment interpolation points and the existing road interpolation points; Obtaining the average angle of the line segment after interpolation of the trajectory road segment interpolation point and the existing road interpolation point; Get the width value and preset weight value of the existing road; The similarity between the trajectory road network and the existing road network data is calculated by using the preset weight value, the average angle of the line segment, the target distance and the width value to obtain a similarity value.

4. The method according to claim 2, characterized in that The step of calculating the confidence of the initial newly added roads and selecting target newly added roads from the initial newly added roads comprises: Performing equidistant interpolation on the initially added roads on the regular grid to obtain a plurality of interpolation points; A window with a preset side length is defined with the pixel of each interpolation point as the center to obtain multiple windows; Count the total pixel flow in each window; Obtaining the number of trajectories of the user signaling trajectory passing through the regular grid to obtain a pixel flow value; Calculate the cross-sectional flow value of the initial newly added road by using the pixel flow sum and the pixel flow value; Calculate the median of the cross-sectional flow value to obtain the newly added road flow value; Get the historical newly added road flow value; Calculating the flow value change by using the historical newly added road flow value and the newly added road flow value; The initial newly added roads corresponding to the flow value change exceeding the preset change value are screened as target newly added roads.

5. The method according to claim 1, characterized in that The step of connecting the existing roads with the target newly added roads and the track road network to complete the automatic road network update includes: Mapping the target newly added road, the removed roads in the trajectory road network, and the existing roads onto a preset map sheet to obtain target newly added road pixels, removed road pixels, and existing road pixels; Convert the target newly added road pixels, the removed road pixels and the existing road pixels into vector lines to obtain target newly added lines, removed lines and reference lines; Using a depth-first traversal strategy to traverse the target newly added lines, the removed lines, and the reference lines to obtain an initial road cluster; Performing anomaly detection and pruning on the initial road clusters to obtain target road clusters; The target newly added lines and removed lines in the target road cluster are attached to the existing roads to complete the automatic update of the road network.

6. The method according to claim 5, characterized in that The step of performing anomaly detection and pruning on the initial road clusters to obtain target road clusters comprises: Eliminate line segments with overlapping endpoints and overlapping lines in the initial road cluster to obtain a reference road cluster; Performing equidistant interpolation on the target newly added lines, the removed lines and the reference lines in the reference road cluster to obtain interpolation points, and obtaining angle values ​​of the road sections corresponding to the interpolation points and the distances between the interpolation points; Abnormal roads in the angle values ​​and the distances that do not conform to preset road rules are pruned to obtain target road clusters.

7. A road network automatic updating device based on signaling data, characterized in that: The device comprises: An acquisition module, used to acquire a user signaling trajectory and map the user signaling trajectory onto a regular grid to obtain a preset feature composite image; An extraction module, used to extract the road centerline from the preset feature composite image using a deep learning model to obtain a trajectory road network; A screening module, used for screening roads based on the trajectory road network to obtain target newly added roads; The attachment module is used to attach the existing roads through the target newly added roads and the track road network to complete the automatic update of the road network.

8. A road network automatic updating device based on signaling data, characterized in that: The device comprises: a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the method for automatic road network update based on signaling data as described in any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the method for automatic road network update based on signaling data as described in any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the method for automatic road network update based on signaling data as claimed in any one of claims 1 to 6 are implemented.

Images