CN115752486A

CN115752486A - Method, device, equipment and medium for acquiring lane line topology based on crowdsourcing track

Info

Publication number: CN115752486A
Application number: CN202211520624.7A
Authority: CN
Inventors: 侯晓辉
Original assignee: Chongqing Changan Automobile Co Ltd
Current assignee: Chongqing Changan Automobile Co Ltd
Priority date: 2022-11-29
Filing date: 2022-11-29
Publication date: 2023-03-07

Abstract

The invention provides a method, a device, equipment and a medium for acquiring lane line topology based on crowdsourcing track, wherein the method comprises the following steps: acquiring track point information collected by crowdsourced vehicles; identifying intersection areas and non-intersection areas based on the track point information; respectively grouping a plurality of track points of the intersection region and the non-intersection region according to the time stamp of the track points and the distance interval of the track points to obtain a track group of the intersection region and a track group of the non-intersection region; clustering the track groups of the intersection area to generate a virtual lane line of the intersection area; clustering the track groups of the non-intersection area to generate lane line topology of the non-intersection area; and merging the virtual lane line of the intersection area and the lane line topology of the non-intersection area to generate a final lane line topological graph. The method generates less lane line redundancy by crowdsourcing trajectory data and utilizing a segmentation, clustering and splicing method, and is closer to reality.

Description

Method, device, equipment and medium for acquiring lane line topology based on crowdsourcing track

Technical Field

The application relates to the technical field of crowdsourcing map data processing, in particular to a method, a device, equipment and a medium for acquiring lane line topology based on crowdsourcing tracks.

Background

In the field of automatic driving, a lane-level high-precision map is important for functions such as path calculation, navigation and positioning. Although the accuracy of the high-precision maps collected professionally is high, the updating frequency often cannot keep up with the change of roads due to the problems of production cost and efficiency.

With the rapid development of sensor technology, wireless communication and network technology, a great deal of large data of space-time trajectories is generated when people go out, and abundant fine road information and human behavior and activity information are contained. The acquisition of the trajectory data is gradually performed by professional department measuring vehicles or professionals, and is changed into a mode of recording the travel trajectory of a non-professional person in a free and voluntary way, and the acquisition of the data is started to be changed into a crowdsourcing mode. The vehicle-mounted trajectory data (crowd-sourced data) in the crowd-sourced mode is undoubtedly the best data source that can provide lane-level road information extraction at present. However, the lane lines generated by using the crowdsourcing trajectories have a problem of high redundancy, and have a certain difference from the actual roads.

Disclosure of Invention

In view of the above drawbacks of the prior art, the present invention provides a method, an apparatus, a device, and a medium for obtaining a lane line topology based on a crowdsourcing trajectory, so as to solve the above technical problems.

The invention provides a method for acquiring lane line topology based on crowdsourcing track, which comprises the following steps:

acquiring track point information acquired by crowdsourcing vehicles, wherein the track point information comprises timestamps and position information of a plurality of track points;

identifying an intersection area and a non-intersection area based on the track point information;

respectively grouping a plurality of track points of the intersection region and the non-intersection region according to the time stamp of the track points and the distance interval of the track points to obtain a track group of the intersection region and a track group of the non-intersection region;

performing distance clustering on the track group of the intersection area to generate a virtual lane line of the intersection area;

performing distance clustering on the track groups of the non-intersection area to generate lane line topology of the non-intersection area;

and merging the virtual lane line of the intersection area and the lane line topology of the non-intersection area to generate a final lane line topological graph.

In an embodiment of the present invention, identifying the intersection area and the non-intersection area based on the information of the track point includes:

calculating the angular speed of the track points according to the timestamps and the direction angles of the track points;

taking the track points with the angular speed larger than a first threshold value as turning candidate points;

combining the continuous turning candidate points into groups, further combining the two groups of non-turning candidate points if the number of the non-turning candidate points between the two groups is less than a second threshold value, and marking the non-turning candidate points between the two groups as turning candidate points;

calculating the direction angle difference and the time stamp difference of the starting point and the end point in each group of turning candidate points, filtering out groups with the difference smaller than a third threshold value, and obtaining a turning point combination;

and performing mean shift (MeanShift) clustering on the turning points in the turning point combination, and identifying the intersection region according to the speed and/or the turning point ratio of the track points.

In an embodiment of the present invention, performing mean shift clustering on the turning points in the turning point combination, and identifying the intersection area according to the speed and/or the proportion of the turning points of the track point includes: carrying out mean shift clustering on the turning points in the turning point combination, wherein the clustering center point is the intersection center point; constructing a plurality of circular rings according to a preset radius and intervals by taking the center point of the intersection as a circle center; calculating the average speed of the track points and the ratio of the turning points to the number of the track points in the annular range from small to large; and if the average speed is greater than a fourth threshold value, or the proportion of the turning points is lower than a fifth threshold value, or the radius of the circular ring exceeds a sixth threshold value, stopping calculation, wherein the circular area between the intersection center point and the circular ring is the intersection area.

In an embodiment of the present invention, grouping the trace points in the intersection region and the non-intersection region according to the timestamp of the trace point and the distance interval of the trace point includes: traversing the trace points according to the time stamp sequence to determine break points; and grouping the track points according to the area where the track points are located and the interruption points to obtain a track group of the intersection area and a track group of the non-intersection area.

In an embodiment of the present invention, the interruption point at least satisfies one of the following conditions: track points at the critical part of the intersection area and the non-intersection area; the time stamp interval of the adjacent track points exceeds a preset threshold value; and the distance interval between the adjacent track points exceeds a preset threshold value.

In an embodiment of the present invention, distance clustering is performed on the trajectory groups of the intersection region to generate a virtual lane line of the intersection region, including:

calculating Hausdorff (Hausdorff) distances of every two track groups of the intersection region, and if the distances of the head and the tail are smaller than a lane width threshold value, taking the smaller distance as the Hausdorff distance of the two track groups; otherwise, taking the distance between the middle parts of the two track groups as a Hausdorff distance;

clustering the track groups of the intersection region according to the Hausdorff distance;

selecting a smoother track group as a representative track from the clustered track clusters to generate a virtual lane line; the calculation method of the smoothness degree is an average value of direction angle difference values of adjacent track points, and the smaller the value is, the smoother the value is.

In an embodiment of the present invention, clustering the trajectory groups of the non-intersection region to generate a lane center line topology of the non-intersection region includes:

clustering the track groups of the non-intersection region according to the passing direction and the distance to obtain road sections;

and clustering the track groups in each road section based on the distance by taking the road sections as units to generate the lane line topology of the non-intersection area.

In an embodiment of the present invention, clustering the trajectory groups of the non-intersection region according to the passing direction and the distance includes:

calculating the distance between the starting point and the end point of each two track groups, taking the smaller value of the two distances as the distance between the track groups, and then carrying out primary clustering on each track group according to the distance;

calculating the Hausdorff distance between the road sections after the primary clustering, and merging the two road sections if the Hausdorff distance is smaller than the lane width; otherwise, calculating the minimum distance between the track groups in the road section, and merging the two road sections if the minimum distance is smaller than the lane width.

In an embodiment of the present invention, if the road segment exceeds 100 meters, the preliminary clustering further includes cutting the road segment into small segments, and the cutting the road segment into small segments includes:

finding track points of the separation positions in the longest track group in the road section according to a preset distance to serve as cutting points; and traversing track points in the track group in the road section, and segmenting the track group according to the distance between the track point and the cutting point.

In an embodiment of the present invention, with the road segments as units, clustering the track groups in each of the road segments based on distance to generate a lane line topology of a non-intersection area, includes:

clustering based on Hausdorff distance is carried out by taking the track group in each road section as an object, and one track group close to the central position of each cluster is screened out from a plurality of track groups of each cluster to be used as a candidate lane line;

generating a map by taking the candidate lane lines as points and the Hausdorff distance between the candidate lane lines as sides, and finding out mutually non-intersected lane lines by utilizing a maximum group algorithm to serve as main lane lines;

and generating a diverging/converging lane line based on the track group data in each road section and the main lane line.

In an embodiment of the present invention, generating a diverging/merging lane line based on the trajectory group data in each of the lane sections and the main lane line includes:

in the road section, acquiring the number of lanes at corresponding positions on track points at the cross section of the road section at intervals by a Gaussian mixture model;

recognizing lanes except the main lane line as candidate diverging/converging lanes by combining the number of lanes at each position and the information of the main lane line;

respectively calculating the minimum distance between the starting point of each candidate branch and confluence lane line and each main lane line, the minimum distance between the end point of each candidate branch and confluence lane line and each main lane line, and the number of main lane lines intersected with each candidate branch and confluence lane line, and judging lane change lane lines according to the calculation result;

filtering out the intersection point of the alternate branching and merging lane line after the lane changing lane line and the main lane line, namely a branching and merging point; if the distance between the starting point of the candidate diverging/converging lane line and the main lane line is not less than the lane width threshold, the track between the starting point of the candidate diverging/converging lane line and the diverging/converging point is a diverging/converging lane line; and if the distance between the end point of the candidate branch and merge lane line and the main lane is not less than the lane width threshold, the track between the end point of the candidate branch and merge lane line and the branch and merge point is a branch and merge lane line.

In an embodiment of the present invention, merging the virtual lane of the intersection region and the lane line topology of the non-intersection region to generate a final lane line topology map, includes:

generating nodes (nodes) at the start and end points of each lane line;

breaking the main lane line at the diverging and converging position to add a node point;

merging the node points with the distances close to the threshold value;

and if one node point is only connected with two lane lines, deleting the node point, merging the lane lines and generating a final lane line topological graph.

The invention provides a device for acquiring lane line topology based on crowdsourcing track, which comprises: the system comprises a crowdsourcing data acquisition module, an information identification module, a grouping module, a cross region processing module, a non-cross region processing module and a fusion module. The crowdsourcing data acquisition module is configured to acquire track point information acquired by crowdsourcing vehicles, and the track point information comprises timestamps and position information of a plurality of track points; the information identification module is configured to identify an intersection area and a non-intersection area based on the information of the track points; the grouping module is configured to respectively group a plurality of track points of the intersection region and the non-intersection region according to the time stamp of the track points and the distance interval of the track points so as to obtain a track group of the intersection region and a track group of the non-intersection region; the intersection region processing module is configured to perform distance clustering on the trajectory groups of the intersection region to generate a virtual lane line of the intersection region; the non-intersection region processing module is configured to distance cluster the trajectory groups of the non-intersection region to generate a lane line topology of the non-intersection region; the fusion module is configured to merge the virtual lane of the intersection region and the lane line topology of the non-intersection region to generate a final lane line topology map.

The electronic device provided by the invention comprises one or more processors and a storage device, wherein the storage device is used for storing one or more programs, and when the one or more programs are executed by the one or more processors, the electronic device is enabled to realize the method for acquiring the lane line topology based on the crowdsourcing track.

The present invention provides a computer-readable storage medium having stored thereon a computer program, which, when executed by a processor of a computer, causes the computer to execute the method for obtaining a lane line topology based on a crowdsourced trajectory according to the present invention.

The invention has the beneficial effects that: the method utilizes the track data acquired by crowdsourced vehicles as input tracks and generates the road-level network topology through the methods of segmentation, aggregation and splicing. The lane line generated by the method has less redundancy and is closer to a real lane.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.

Drawings

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 application. It is obvious that the drawings in the following description are only some embodiments of the application, and that for a person skilled in the art, other drawings can be derived from them without inventive effort. In the drawings:

fig. 1 is a schematic diagram of an implementation environment for acquiring a lane line topology based on a crowd-sourced trajectory according to an exemplary embodiment of the present application;

FIG. 2 is a flow diagram illustrating a method of obtaining lane line topology based on crowd-sourced trajectories in an exemplary embodiment of the application;

FIG. 3 is a flow diagram illustrating the identification of intersection and non-intersection areas based on the track point information in accordance with an exemplary embodiment of the present application;

fig. 4 is a flowchart illustrating MeanShift clustering of turning points in the turning point group and identifying the intersection area according to the speed of the track point and/or the turning point according to an exemplary embodiment of the present application;

FIG. 5 is a flow chart illustrating grouping of trace points of the intersection region and the non-intersection region according to a timestamp of the trace point and a distance separation of the trace point, respectively, according to an exemplary embodiment of the present application;

FIG. 6 is a flow chart illustrating clustering of groups of trajectories of the intersection region to generate a virtual lane line for the intersection region in accordance with an exemplary embodiment of the present application;

FIG. 7 is a schematic illustration of a virtual lane line of a crosshatch area as shown in an exemplary embodiment of the present application;

FIG. 8 is a flow chart illustrating clustering of the trajectory groups of the non-intersection regions to generate a lane line topology for the non-intersection regions in accordance with an exemplary embodiment of the present application;

FIG. 9 is a flow chart illustrating clustering of trajectory groups of the non-intersection regions based on traffic direction and distance according to an exemplary embodiment of the present application;

FIG. 10 is a flow chart illustrating cutting a longer road segment in an exemplary embodiment of the present application;

FIG. 11 is a flow chart illustrating distance-based clustering of groups of tracks in each of the road segments in units of the road segment according to an exemplary embodiment of the present application;

FIG. 12 is a flow chart illustrating the generation of diverging merge lane lines based on trajectory set data in each of the road segments and the primary lane lines in accordance with an exemplary embodiment of the present application;

FIG. 13 is a schematic diagram of a main lane line and a bifurcation lane line configuration shown in an exemplary embodiment of the present application;

FIG. 14 is a schematic diagram of the structure of a main lane line and a merge lane line as shown in an exemplary embodiment of the present application;

FIG. 15 is a flowchart illustrating merging a virtual lane line of the intersection region and a lane line topology of the non-intersection region to generate a final lane line topology map in accordance with an exemplary embodiment of the present application;

fig. 16 is a block diagram illustrating an apparatus for obtaining lane line topology based on crowd-sourced trajectories according to an exemplary embodiment of the present application;

fig. 17 is a schematic structural diagram illustrating a computer system suitable for implementing an electronic device according to an exemplary embodiment of the present application.

Detailed Description

Other advantages and effects of the present invention will become apparent to those skilled in the art from the disclosure herein, wherein the embodiments of the present invention are described in detail with reference to the accompanying drawings and preferred embodiments. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It should be understood that the preferred embodiments are illustrative of the invention only and are not limiting upon the scope of the invention.

It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention, and the components related to the present invention are only shown in the drawings rather than drawn according to the number, shape and size of the components in actual implementation, and the type, quantity and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated.

In the following description, numerous details are set forth to provide a more thorough explanation of embodiments of the present invention, however, it will be apparent to one skilled in the art that embodiments of the present invention may be practiced without these specific details, and in other embodiments, well-known structures and devices are shown in block diagram form, rather than in detail, in order to avoid obscuring embodiments of the present invention.

First, the crowdsourcing trajectory is trajectory data collected by crowdsourcing vehicles, and the trajectory data collected by the crowdsourcing vehicles is space-time position information fed back by the crowdsourcing vehicles in a crowdsourcing manner through intelligent terminal devices (such as vehicle-mounted navigation devices, smart phones, vehicle-mounted computers, tablet computers, and notebook computers).

The lane line topology is topology information that can reflect the relative positions of lane lines and lane lines in a road.

The time stamp is data generated by using a digital signature technology, and a signed object comprises original file information, signature parameters, signature time and other information. The main purpose of the time stamp is to authenticate the time of data generation by a certain technical means, so as to verify whether the data is falsified after being generated.

Mean shift (Mean shift) clustering, as the name implies, consisting of Mean and shift, is a sliding window based algorithm that attempts to find areas where data points are dense.

And calculating a drift vector of the central point through the data density change in the region of interest, so as to move the central point for the next iteration until the position with the maximum density is reached (the central point is unchanged). This can be done starting from each data point and counting the number of times data appear in the region of interest, which will ultimately be the basis for classification.

The Hausdorff distance is a measure describing the degree of similarity between two sets of points, and is a defined form of the distance between two sets of points: assuming that there are two sets of a = { a1 …, ap), B = { B1 …, bq }, the Hausdorff distance between the two sets of points is defined as h (a, B) = max a ∈ a { minb ∈ B { d (a, B) }

Where a and B are points in the set a and the set B, respectively, and d (a and B) represents the Euclidean distance between a and B. From the above formula, the Hausdorff distance h (a, B) measures the maximum degree of mismatch between two point sets, and a smaller distance indicates a higher degree of match.

The gaussian mixture model is a model that accurately quantifies objects by using a gaussian probability density function (normal distribution curve), and is formed by decomposing objects into a plurality of objects based on the gaussian probability density function (normal distribution curve).

Fig. 1 is a schematic diagram of an implementation environment for acquiring a lane line topology based on a crowd-sourced trajectory according to an exemplary embodiment of the present application. As shown in fig. 1, in the driving process of the vehicle, navigation is implemented through the navigation map software installed on the intelligent terminal 110, and the navigation map software refreshes the road condition, that is, the navigation map software makes a network request to the navigation server 120 according to the domain name of the navigation server 120, and then the navigation server 120 returns a corresponding navigation path to the navigation map software. The intelligent terminal 110 shown in fig. 1 may be a terminal device supporting installation of navigation map software, such as a smart phone, a vehicle-mounted computer, a tablet computer, and the like, but is not limited thereto. The navigation server 120 is a navigation server, and may be, for example, an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a Network service, cloud communication, a middleware service, a domain name service, a security service, a CDN (Content Delivery Network), a big data and an artificial intelligence platform, which is not limited herein. The intelligent terminal 110 may communicate with the navigation server 220 through a wireless network such as 3G (third generation mobile information technology), 4G (fourth generation mobile information technology), 5G (fifth generation mobile information technology), etc., which is not limited herein.

In the field of automatic driving, high-precision maps at lane level are mostly adopted, and although the high-precision maps have higher precision, the updating frequency cannot keep up with the change of roads due to the problems of production cost and efficiency. To solve the problems, embodiments of the present application respectively provide a method for obtaining a lane line topology based on a crowdsourcing trajectory, an apparatus for obtaining a lane line topology based on a crowdsourcing trajectory, an electronic device, a computer-readable storage medium, and a computer program product, which will be described in detail below.

Referring to fig. 2, fig. 2 is a flowchart illustrating a method for obtaining a lane line topology based on a crowd-sourced trajectory according to an exemplary embodiment of the present disclosure. The method may be applied to the implementation environment shown in fig. 1 and specifically executed by the intelligent terminal 110 in the implementation environment. It should be understood that the method may be applied to other exemplary implementation environments and is specifically executed by devices in other implementation environments, and the embodiment does not limit the implementation environment to which the method is applied.

As shown in fig. 2, in an exemplary embodiment, the method for obtaining the lane line topology based on the crowdsourcing trajectory at least includes steps S210 to S260, which are described in detail as follows:

and step S210, acquiring track point information collected by crowdsourced vehicles.

The navigation server 120 may obtain crowd-sourced trajectory data of a corresponding road from crowd-sourced vehicles, and the crowd-sourced vehicles acquire the trajectory data through vehicle-end sensors when driving on the road, where the crowd-sourced trajectory data is obtained through a crowd-sourced mode. The vehicle-end sensor may include a Global Navigation Satellite System (GNSS), a camera, an ultrasonic radar, a laser radar, or a millimeter-wave radar, and the like, and the corresponding crowd-sourced trajectory data may include a timestamp and position information of the trajectory point, such as a direction angle, and the like.

And step 220, identifying the intersection area and the non-intersection area based on the track point information.

The road usually comprises an intersection road section and a non-intersection road section, and the vehicle tracks of the intersection road section and the non-intersection road section are different, so that after track points acquired by crowdsourced vehicles are acquired, an intersection area and a non-intersection area are firstly identified according to track point information for subsequent track segmentation.

Fig. 3 is a flow chart of step S220 in the embodiment shown in fig. 2 in an exemplary embodiment. As shown in fig. 3, identifying intersection areas and non-intersection areas based on the track point information may include steps S310 to S350, which are described in detail as follows:

and S310, calculating the angular speed of the track point according to the timestamp and the direction angle of the track point.

The track point information obtained in step S210 includes a timestamp and a direction angle of each track point, and an angular velocity of each track point is respectively calculated according to a formula ω = Δ θ/Δ t according to the timestamp and the direction corresponding to each track point, where ω is an angular velocity, Δ t represents time, and Δ θ represents an angle turned within Δ t time.

And step S320, taking the track points with the angular speed larger than the first threshold value as turning candidate points.

And (4) recording the angular velocity corresponding to each track point calculated in the step (S310) and a preset threshold as a first threshold for comparison, if the angular velocity corresponding to the track point is greater than the first threshold, marking the track point as a turning candidate point, otherwise, marking the track point as a non-turning candidate point. It should be noted that the selection of the first threshold is related to the timestamp interval of the trace point, and the like, and is not limited herein, and may be specifically set according to actual requirements. The same is true for the thresholds below.

Step S330, combining the continuous steering candidate points into a group; and if the number of the non-turning candidate points between the two groups is less than a second threshold value, further combining the two groups, and marking the non-turning candidate points between the two groups as turning candidate points.

And (3) dividing a plurality of track points into turning candidate points and non-turning candidate points by calculating the angular speed of each track point and comparing the angular speed with a preset threshold, and combining the successive track points into groups if the successive track points are the turning candidate points, so as to finish combining the successive turning candidate points. And comparing the steering candidate point groups pairwise, if the number of the non-steering candidate points between the two steering candidate points is smaller than a preset threshold, and the number is marked as a second threshold, further combining the two steering candidate points, and marking the non-steering candidate points between the two steering candidate points as steering candidate points.

Step S340, calculating a direction angle difference and a timestamp difference between the starting point and the ending point in each group of turning candidate points, and filtering out groups with differences smaller than a third threshold value to obtain a turning point group.

And respectively calculating a direction angle difference value and a time stamp difference value of the starting point and the end point in each group of turning candidate points, and comparing the direction angle difference value and the time stamp difference value with a preset threshold value, and recording the direction angle difference value and the time stamp difference value as a third threshold value, wherein the third threshold value comprises a direction angle difference value threshold value and a time stamp difference value threshold value, the direction angle difference value threshold value is 50 degrees for example, and the time stamp difference value threshold value is 3 seconds for example. If the direction angle difference value of the starting point and the end point of the turning candidate point group is smaller than the direction angle difference value threshold value and/or the time stamp difference value of the starting point and the end point is smaller than the time stamp difference value threshold value, the turning candidate point group is filtered, and the rest turning candidate point group is the turning point group.

And S350, performing mean shift clustering on the turning points in the turning point group, and identifying the intersection area according to the speed of the track point and/or the turning point.

Fig. 4 is a flow chart of step S350 in the embodiment shown in fig. 3 in an exemplary embodiment. As shown in fig. 4, the process of performing MeanShift clustering on the turning points in the turning point group and identifying the intersection area according to the speed and/or turning points of the track point may include steps S410 to S430, which are described in detail as follows:

and S410, carrying out MeanShift clustering (selecting 50 meters in bandwidth) on the filtered steering points, wherein the clustering center point is the intersection center point.

Step S420, constructing a plurality of circular rings according to a preset radius and intervals by taking the center point of the intersection as a circle center;

the preset radius and the interval can be set according to actual conditions, for example, a plurality of concentric rings are sequentially constructed from small to large by starting with a radius of 30 meters and taking 5 meters as the interval between the rings.

And S430, calculating the average speed of the track points in the annular range and the ratio of the turning points to the number of the track points from small to large, and identifying the intersection region according to the calculation result.

Specifically, the circular range is calculated from small to large, for example, the average velocity V = (V1 + V2+ V3+ … … + Vn)/n of all track points in the circular range between the circumference with the radius of 30 meters and the circumference with the radius of 35 meters is calculated first, where Vn represents the velocity of the nth track point, and n represents the number of track points; and the ratio of the number of turning points to the total number of track points in the circular range. And then sequentially calculating the average speed and the steering point ratio of all track points in the annular range between the circumference with the radius of 35 meters and the circumference with the radius of 40 meters in the same way, and repeating the steps until the calculation is stopped. If the average speed V is larger than a preset threshold, recording as a fourth threshold, or recording as a fifth threshold if the ratio of the number of the turning points to the number of the track points is lower than the preset threshold, or recording as a sixth threshold if the radius of the circular ring exceeds the preset threshold, and stopping calculation, wherein at this moment, the circular area between the intersection central point and the circular ring is the intersection area. The area outside the intersection area is a non-intersection area.

Referring to fig. 2, in step S230, a plurality of trace points in the intersection area and the non-intersection area are grouped according to the timestamp of the trace point and the distance interval of the trace point, so as to obtain a trace group in the intersection area and a trace group in the non-intersection area.

Fig. 5 is a flow chart of step S230 in the embodiment shown in fig. 2 in an exemplary embodiment. As shown in fig. 5, the process of grouping the trace points of the intersection region and the non-intersection region according to the timestamp of the trace point and the distance interval of the trace point may include steps S510 to S520, which are described in detail as follows:

and step S510, traversing the track points according to the sequence of the timestamps, and determining the interruption points.

The trace point information comprises timestamps, the trace points are traversed according to the sequence of the timestamps, and break points are selected from the trace points, wherein the break points at least meet the following conditions: (1) Track points at the critical position of an intersection area and a non-intersection area; (2) The time stamp interval of adjacent track points exceeds a preset threshold value; and (3) the distance interval between adjacent track points exceeds a preset threshold value.

And S520, grouping the track points according to the areas where the track points are located and the interruption points.

The area where the track points are located comprises a crossing area and a non-crossing area, and a plurality of track points in the crossing area are divided into track groups of a plurality of crossing areas according to the interruption points; the plurality of trace points in the non-intersection region are divided into a plurality of trace groups of non-intersection regions.

Continuing with fig. 2, step S240, the trajectory groups of the intersection region are clustered to generate a virtual lane line of the intersection region. The intersection region processing is aimed at obtaining a virtual connecting lane, that is, a connection relationship between lanes on each road associated with the intersection.

Referring to fig. 6, fig. 6 is a flowchart of step S240 in the embodiment shown in fig. 2 in an exemplary embodiment. As shown in fig. 6, the process of clustering the trajectory groups of the intersection region to generate the virtual lane lines of the intersection region may include steps S610 to S630, which are described in detail as follows:

and step S610, calculating the Hausdorff distance of each two track groups of the intersection region.

The Hausdorff distance in the embodiment is an improved Hausdorff distance, specifically, firstly, the Hausdorff distance of the head and tail parts of the two track groups is calculated, and if the distance of the head and tail parts is smaller than a lane width threshold, the distance is used as the Hausdorff distance of the two track groups; otherwise, the distance between the middle parts of the two tracks is calculated according to the standard Hausdorff distance. The improved method is adopted in the embodiment because the standard Hausdorff distance is the distance between the middle parts of the two track groups, if the head and tail parts of the two track groups are close to each other, the middle parts have larger distance, and the two track groups can be divided into different clusters according to the distance clustering. However, for the virtual lane at the intersection, if the two tracks approach from beginning to end, the two tracks are the same in the starting lane and the ending lane, and should belong to the same virtual lane. Based on this, the present embodiment improves the Hausdorff distance, preventing the above-described problems from occurring.

And step 620, clustering the track group of the intersection region according to the Hausdorff distance.

Distance clustering is performed according to the Hausdorff distance calculated in step S610.

And step S630, selecting a smoother track group as a representative track from the clustered track clusters to generate a virtual lane line.

And when the representative track is selected, the steering smoothness of the track is taken as the weight, wherein the calculation method of the smoothness is that the average value of the direction angle difference values of the adjacent track points in the track group is calculated. Smaller averages represent smoother. Referring to fig. 7, fig. 7 is a virtual lane line generated in step S240 in the embodiment shown in fig. 2.

Continuing to refer to fig. 2, in step S250, the trajectory groups of the non-intersection area are clustered to generate the lane line topology of the non-intersection area.

The non-intersection area processing aims to obtain the topology of the lane lines, including a diverging lane line and a converging lane line.

Referring to fig. 8, fig. 8 is a flowchart of step S250 in the embodiment shown in fig. 2 in an exemplary embodiment. As shown in fig. 8, the process of clustering the trajectory groups of the non-intersection region to generate the lane line topology of the non-intersection region may include steps S810 to S820, which are described in detail as follows:

and step 810, clustering the track groups of the non-intersection areas according to the passing direction and the distance to obtain road sections.

Namely, the track group which is positioned between two intersections and has the same passing direction is classified into a road section.

Referring to fig. 9, fig. 9 is a flowchart of step S810 in the embodiment shown in fig. 8 in an exemplary embodiment. As shown in fig. 9, the process of clustering the trajectory groups of the non-intersection region according to the passing direction and the distance may include steps S910 to S920, which are described in detail as follows:

step S910, calculating the distance between the starting point and the end point of each two track groups, taking the smaller value of the two distances as the distance between the track groups, and then carrying out primary clustering (hierarchical clustering) on each track group according to the distance.

Considering that the track groups are discontinuous or incomplete, only considering that the distance between the starting point and the end point can cause the wrong road segment allocation of the isolated track groups, further optimization is needed. The optimization step is S920.

And step S920, calculating the Hausdorff distance between the preliminarily clustered road segments.

If the Hausdorff distance between the road sections is smaller than the lane width, merging the two road sections; otherwise, calculating the minimum distance between the track groups in the road section, and merging the two road sections if the minimum distance is smaller than the lane width.

Due to the existence of lane change in reality, if the road segment generated in step S810 is too long, the distance between the track groups in the road segment changes frequently, resulting in poor clustering effect. To address this problem, road segments of more than 100 meters are cut prior to clustering. The cutting steps are as follows:

and step S1010, finding track points of the separation positions in the track group according to a preset distance to serve as cutting points.

The preset interval is less than 100 meters, and can be set according to actual conditions, so that the long road section with the length of more than 100 meters is divided into small sections.

Step S1020, traversing the track points in the track group, and segmenting the track group according to the distance between the track point and the cutting point.

And the track points between the adjacent cutting points are a small road section.

Referring to fig. 8, in step S820, distance-based clustering is performed on the track group in each road segment by taking the road segment as a unit, so as to generate a lane line topology.

It should be noted that the lane lines include a main lane line, a branch lane line, and a merge lane line.

Referring to fig. 11, fig. 11 is a flowchart of step S820 in the embodiment shown in fig. 8 in an exemplary embodiment. As shown in fig. 11, the process of clustering the track group in each road segment based on distance by using the road segment as a unit to generate the lane line topology may include steps S1110 to S1130, which are described in detail as follows:

step S1110, clustering based on the Hausdorff distance is performed with the trajectory group in each road segment as an object, and one trajectory group closer to the center position of each cluster is selected from the plurality of trajectory groups of each cluster as a candidate lane line.

Step S1120, clustering is performed with the candidate lane lines as objects, and a main lane line is generated.

The specific process is as follows: generating a map by taking the candidate lane lines as points and taking Hausdorff distance between the candidate lane lines as sides, and finding out mutually disjoint lane lines by utilizing a maximum clique algorithm (max _ weight _ clique) to be used as a main lane line; in the maximum clique algorithm, the smoother lane line is preferentially selected in consideration of the smoothness degree weight of the lane line.

Step S1130 generates a diverging/converging lane line based on the trajectory group data in each of the road segments and the main lane line. Referring to fig. 12, the method specifically includes steps S1210 to S1240:

and S1210, in the road section, acquiring the number of lanes at the corresponding position on the track points at the cross section of the road section at certain intervals by a Gaussian mixture model.

The distance of the gap is not limited herein, and may be set according to specific situations, for example, the distance of the gap is 20 meters or 30 meters, and it should be understood that the distance of the gap is less than the length of the road section.

And step S1220, combining the number of lanes at each position and the information of the main lane line, and identifying lanes other than the main lane line as candidate diverging/converging lane lines.

And step S1230, determining a lane change lane from the candidate diverging/converging lane line and the main lane line.

The method comprises the following specific steps: the minimum distance min _ dis _ s between the start point of each candidate diverging/merging lane line and each main lane line, the minimum distance min _ dis _ e between the end point of each candidate diverging/merging lane line and each main lane, and the number inter _ cnt of main lane lines intersecting each candidate diverging/merging lane line are calculated, respectively. A candidate diverging/merging lane satisfying one of the following conditions, that is, a lane change lane: (1) max (min _ dis _ s, min _ dis _ e) < lane width threshold; inter _ cnt >1.

And step S1240, finding out the branch and merge lane line based on the candidate branch and merge lane line, the main lane line and the lane change lane line.

The specific process is as follows: filtering out the intersection point of the alternate branching and merging lane line after the lane changing lane line and the main lane line, namely a branching and merging point; generating a section of track by the starting point of the candidate diverging and converging lane and the diverging and converging point, and marking the section of track as s _ lane, and generating a section of track by the end point of the candidate diverging and converging lane and the diverging and converging point, and marking the section of track as e _ lane; if the distance between the starting point of the candidate diverging/converging lane and the main lane is not less than the lane width threshold, reserving s _ lane as a diverging/converging lane line; and if the distance between the end point of the candidate diverging/converging lane and the main lane is not less than the lane width threshold, reserving e _ lane as a diverging/converging lane line. Examples of diverging and merging lane lines are shown in fig. 13 and 14, where fig. 13 is a schematic view of a main lane and diverging lanes (the dashed lines are diverging lanes), and fig. 14 is a schematic view of a main lane and merging lanes (the dashed lines are merging lanes).

Continuing to refer to fig. 2, step S260 is executed to combine the virtual lane line of the intersection region and the lane line topology of the non-intersection region, so as to generate a final lane line topology map.

Referring to fig. 15, fig. 15 is a flowchart of step S260 in the embodiment shown in fig. 2 in an exemplary embodiment. As shown in fig. 15, the process of merging the virtual lane line of the intersection region and the lane line topology of the non-intersection region to generate the final lane line topology map may include steps S1510 to S1530, which are described in detail as follows:

in step S1510, node points are generated at the start point and the end point of each lane line.

Each lane comprises a virtual lane line in an intersection area, a main lane line in a non-intersection area, and a diverging and converging lane line. The start and end points of the respective lanes are marked as node points.

Step S1520, break the main lane line at the branch and merge position to add a node point.

The main lane line is broken at the intersection of the diverging lane line and the main lane line and at the intersection of the converging lane line and the main lane line to add a node point.

And step S1530, merging the node points with the distances close to the threshold value.

The threshold value here is not limited, and may be set according to actual conditions. And if one node point is only connected with two lane lines, deleting the node point, merging the lane lines and generating the final lane center line topology.

Fig. 16 is a block diagram illustrating an apparatus for obtaining a lane line topology based on a crowd-sourced trajectory according to an exemplary embodiment of the present application. The device can be applied to the implementation environment shown in fig. 1 and is specifically configured in the intelligent terminal 110. The apparatus may also be applied to other exemplary implementation environments, and is specifically configured in other devices, and the embodiment does not limit the implementation environment to which the apparatus is applied.

As shown in fig. 16, the exemplary apparatus for obtaining a lane line topology based on a crowd-sourced trajectory includes: the crowd-sourced data acquisition module 1601, the information identification module 1602, the grouping module 1603, the intersection area processing module 1604, the non-intersection area processing module 1605 and the fusion module 1606, wherein the crowd-sourced data acquisition module 1601 is configured to acquire track point information acquired by a crowd-sourced vehicle, and the track point information includes timestamps and position information of a plurality of track points; the information identification module 1602 is configured to identify intersection areas and non-intersection areas based on the information of the track points; the grouping module 1603 is configured to group the plurality of track points of the intersection area and the non-intersection area respectively according to the timestamp of the track point and the distance interval of the track point to obtain a track group of the intersection area and a track group of the non-intersection area; the intersection region processing module 1604 is configured to distance cluster the trajectory groups of the intersection region to generate a virtual lane line of the intersection region; the non-intersection region processing module 1605 is configured to perform distance clustering on the trajectory groups of the non-intersection region to generate a lane line topology of the non-intersection region; the fusion module 1606 is configured to merge the virtual lane of the intersection region and the lane line topology of the non-intersection region, generating a final lane line topology map.

It should be noted that the apparatus for obtaining the lane line topology based on the crowdsourcing trajectory provided in the foregoing embodiment and the method for obtaining the lane line topology based on the crowdsourcing trajectory provided in the foregoing embodiment belong to the same concept, and specific ways in which the modules and units perform operations have been described in detail in the method embodiment, and are not described herein again. In practical applications, the road condition refreshing apparatus provided in the above embodiment may distribute the above functions through different functional modules according to needs, that is, divide the internal structure of the apparatus into different functional modules to complete all or part of the above described functions, which is not limited herein.

An embodiment of the present application further provides an electronic device, including: one or more processors; a storage device, configured to store one or more programs, which when executed by the one or more processors, cause the electronic device to implement the method for obtaining a lane line topology based on a crowd-sourced trajectory provided in the above embodiments.

FIG. 17 illustrates a schematic structural diagram of a computer system suitable for use in implementing the electronic device of an embodiment of the present application. It should be noted that the computer system 1700 of the electronic device shown in fig. 17 is only an example, and should not bring any limitation to the functions and the scope of the application of the embodiments.

As shown in fig. 17, a computer system 1700 includes a Central Processing Unit (CPU) 1701 that can perform various appropriate actions and processes, such as executing the methods described in the above embodiments, according to a program stored in a Read-Only Memory (ROM) 1702 or a program loaded from a storage portion 1708 into a Random Access Memory (RAM) 1703. In the RAM 1703, various programs and data necessary for system operation are also stored. The CPU 1701, ROM 1702, and RAM 1703 are connected to each other through a bus 1704. An Input/Output (I/O) interface 1705 is also connected to the bus 1704.

The following components are connected to the I/O interface 1705: an input section 1706 including a keyboard, a mouse, and the like; an output section 1707 including a Display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and a speaker; a storage portion 1708 including a hard disk and the like; and a communication section 1709 including a Network interface card such as a LAN (Local Area Network) card, a modem, or the like. The communication section 1709 performs communication processing via a network such as the internet. A driver 1710 is also connected to the I/O interface 1705 as necessary. A removable medium 1711 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 1710 as necessary, so that a computer program read out therefrom is mounted into the storage portion 1708 as necessary.

In particular, according to embodiments of the application, the processes described above with reference to the flow diagrams may be implemented as computer software programs. For example, embodiments of the present application include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising a computer program for performing the method illustrated by the flow chart. In such embodiments, the computer program may be downloaded and installed from a network via the communication portion 1709, and/or installed from the removable media 1711. When the computer program is executed by a Central Processing Unit (CPU) 1701, various functions defined in the system of the present application are executed.

It should be noted that the computer readable medium shown in the embodiments of the present application may be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may be, for example, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a Read-Only Memory (ROM), an Erasable Programmable Read-Only Memory (EPROM), a flash Memory, an optical fiber, a portable Compact Disc Read-Only Memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer-readable signal medium may comprise a propagated data signal with a computer-readable computer program embodied therein, either in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. The computer program embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the foregoing.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. Each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The units described in the embodiments of the present application may be implemented by software, or may be implemented by hardware, and the described units may also be disposed in a processor. Wherein the names of the elements do not in some way constitute a limitation on the elements themselves.

Another aspect of the present application also provides a computer-readable storage medium, on which a computer program is stored, which, when executed by a processor of a computer, causes the computer to execute the method for obtaining a lane line topology based on a crowd-sourced trajectory as described above. The computer-readable storage medium may be included in the electronic device described in the above embodiment, or may exist separately without being incorporated in the electronic device.

Another aspect of the application also provides a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions to cause the computer device to execute the method for acquiring the lane line topology based on the crowdsourcing trajectory provided in the above embodiments.

The foregoing embodiments are merely illustrative of the principles of the present invention and its efficacy, and are not to be construed as limiting the invention. Those skilled in the art can modify or change the above-described embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.

Claims

1. A method for obtaining lane line topology based on crowd-sourced trajectories, the method comprising:

identifying intersection areas and non-intersection areas based on the track point information;

clustering the track groups of the intersection area to generate a virtual lane line of the intersection area;

clustering the track groups of the non-intersection areas to generate lane line topology of the non-intersection areas;

2. The method of claim 1, wherein identifying intersection and non-intersection areas based on the track point information comprises:

and carrying out mean shift clustering on the turning points in the turning point combination, and identifying the intersection area according to the speed and/or the turning point ratio of the track points.

3. The method of claim 2, wherein performing mean-shift clustering of turning points in the combination of turning points and identifying the intersection area based on the speed and/or turning point ratio of the track points comprises:

performing mean shift clustering on the turning points in the turning point combination, wherein a clustering center point is an intersection center point;

constructing a plurality of circular rings according to a preset radius and intervals by taking the center point of the intersection as a circle center;

calculating the average speed of the track points in the annular range and the ratio of the turning points to the number of the track points from small to large;

and if the average speed is greater than a fourth threshold value, or the proportion of the turning points is lower than a fifth threshold value, or the radius of the circular ring exceeds a sixth threshold value, stopping calculation, wherein the circular area between the intersection center point and the circular ring is the intersection area.

4. The method of claim 1, wherein grouping the plurality of trace points of the intersection region and the non-intersection region according to the time stamps of the trace points and the distance intervals of the trace points, respectively, comprises:

traversing the trace points according to the time stamp sequence to determine break points;

and grouping the track points according to the area where the track points are located and the interruption points to obtain a track group of the intersection area and a track group of the non-intersection area.

5. The method according to claim 4, wherein said break point satisfies at least one of the following conditions:

track points at the critical part of the intersection area and the non-intersection area;

the time stamp interval of the adjacent track points exceeds a preset threshold value;

and the distance interval between the adjacent track points exceeds a preset threshold value.

6. The method of claim 1, wherein clustering the groups of trajectories for the intersection region to generate a virtual lane line for the intersection region comprises:

calculating the Hausdorff distance of every two track groups in the intersection region, and if the distance between the head part and the tail part is smaller than the threshold value of the lane width, taking the smaller distance as the Hausdorff distance of the two track groups; otherwise, taking the distance between the middle parts of the two track groups as a Housdov distance;

and selecting a smoother track group from the clustered track clusters as a representative track to generate a virtual lane line.

7. The method of claim 1, wherein clustering the group of trajectories of the non-intersection region to generate a lane line topology for the non-intersection region comprises:

and clustering the track groups in each road section based on the distance by taking the road section as a unit to generate the lane line topology of the non-intersection area.

8. The method of claim 7, wherein clustering the group of trajectories of the non-intersecting port region according to direction of traffic and distance comprises:

calculating the Hausdorff distance between the road segments after the primary clustering, and merging the two road segments if the Hausdorff distance is smaller than the lane width; otherwise, calculating the minimum distance between the track groups in the road section, and merging the two road sections if the minimum distance is smaller than the lane width.

9. The method of claim 7, wherein if the road segment exceeds 100 meters, the preliminary clustering further comprises cutting the road segment into small segments, the cutting the road segment into small segments comprising:

finding track points of the separation positions in the longest track group in the road section according to a preset distance to serve as cutting points;

and traversing track points in the track group in the road section, and segmenting the track group according to the distance between the track point and the cutting point.

10. The method according to claim 7, wherein performing distance-based clustering on the track group in each road segment in units of the road segment to generate a lane line topology of a non-intersection area comprises:

clustering based on the Hausdorff distance by taking the track group in each road section as an object, and screening out one track group close to the central position of each cluster from a plurality of track groups of each cluster to be used as a candidate lane line;

generating a map by taking the candidate lane lines as points and the Hausdorff distance between the candidate lane lines as sides, and finding out mutually non-intersected lane lines by utilizing a maximum group algorithm to be used as main lane lines;

11. The method of claim 10, wherein generating a diverging merge lane line based on the trajectory set data in each of the lane segments and the primary lane line comprises:

recognizing lanes except the main lane line as candidate diverging/converging lane lines by combining the number of lanes at each position and the information of the main lane line;

filtering out the intersection point of the alternate branching and merging lane line after the lane changing lane line and the main lane line, namely a branching and merging point; if the distance between the starting point of the candidate branch merging lane line and the main lane line is not less than the lane width threshold value, the track between the starting point of the candidate branch merging lane line and the branch merging point is a branch merging lane line; and if the distance between the end point of the candidate branch and merge lane line and the main lane is not less than the lane width threshold, the track between the end point of the candidate branch and merge lane line and the branch and merge point is a branch and merge lane line.

12. The method of claim 10, wherein merging the virtual lanes of the intersection region and the lane line topology of the non-intersection region to generate a final lane line topology map comprises:

generating nodes at the starting point and the ending point of each lane line;

breaking the main lane line at the branching and merging position to add nodes;

merging the nodes with the distances close to the threshold value;

and if one node is only connected with two lane lines, deleting the node, merging the lane lines and generating a final lane line topological graph.

13. An apparatus for obtaining lane line topology based on crowd-sourced trajectories, the apparatus comprising:

the crowdsourcing data acquisition module is configured to acquire track point information acquired by crowdsourcing vehicles, and the track point information comprises timestamps and position information of a plurality of track points;

the information identification module is configured to identify an intersection area and a non-intersection area based on the information of the track points;

the grouping module is configured to respectively group a plurality of track points of the intersection region and the non-intersection region according to the time stamp of the track points and the distance interval of the track points so as to obtain a track group of the intersection region and a track group of the non-intersection region;

the intersection area processing module is configured to perform distance clustering on the track groups of the intersection area so as to generate a virtual lane line of the intersection area;

the non-intersection region processing module is configured to perform distance clustering on the track groups of the non-intersection region to generate lane line topology of the non-intersection region;

and the fusion module is configured to combine the virtual lane of the intersection region and the lane line topology of the non-intersection region to generate a final lane line topological graph.

14. An electronic device, characterized in that the electronic device comprises:

one or more processors;

storage means for storing one or more programs that, when executed by the one or more processors, cause the electronic device to implement the method of obtaining lane line topology based on crowd-sourced trajectories of any one of claims 1-12.

15. A computer-readable storage medium, having stored thereon a computer program which, when executed by a processor of a computer, causes the computer to execute the method of obtaining a lane line topology based on crowd-sourced trajectories of any one of claims 1 to 12.