CN114116854A

CN114116854A - Track data processing method, device, equipment and storage medium

Info

Publication number: CN114116854A
Application number: CN202111502808.6A
Authority: CN
Inventors: 杨远航
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2021-12-09
Filing date: 2021-12-09
Publication date: 2022-03-01

Abstract

The embodiment of the disclosure provides a track data processing method, a track data processing device, track data processing equipment and a computer readable storage medium. The method provided by the embodiment of the disclosure performs data mining based on road network data and vehicle driving track data, extracts a plurality of representative characteristics related to the motor vehicle signal lamp for training of a machine learning model, and accordingly determines the existence of the motor vehicle signal lamp. In addition, the method provided by the embodiment of the disclosure further determines the type of the motor vehicle signal lamp based on the road network data and the track data in the specific driving direction under the condition that the motor vehicle signal lamp is determined to exist. By the method, the accuracy of identifying the motor vehicle signal lamp is improved, the manual operation cost for subsequent verification is saved, and the user experience is remarkably improved based on the traffic scheme provided by more accurate motor vehicle signal lamp identification. The track data processing method disclosed by the invention can be applied to the application fields of maps, intelligent transportation and the like.

Description

Track data processing method, device, equipment and storage medium

Technical Field

The present disclosure relates to the field of data mining, and more particularly, to a method, an apparatus, a device, and a storage medium for processing trajectory data.

Background

With the continuous development of urbanization, the urban traffic problem becomes more and more obvious. The intelligent traffic system is a novel traffic system and aims to effectively solve the existing traffic problems. Data mining utilizes a large amount of data generated by daily traffic behaviors to extract information useful for predicting traffic, and with the development of intelligent traffic, data mining technology is more and more widely applied to the traffic field, and various current traffic scheme providers rely on data mining of traffic information.

Currently, the main research direction of intelligent traffic systems is the control and inducement of traffic flow, which can be achieved by using traffic lights. Traffic lights are one of the most important elements on roads, and traffic plan providers can determine traffic lights on roads based on road network data they acquire. However, due to the limitation of the road network data collection period, the update period of the traffic signal data is also long, and the actual traffic signal may be adjusted frequently according to the city plan, so that the traffic signal data utilized by the traffic plan provider may have a deviation from the current actual situation.

Therefore, an efficient and accurate traffic signal light identification method is needed.

Disclosure of Invention

In order to solve the problems, the method extracts various representative characteristics for machine learning model training by mining the track data and road network data of vehicle driving, thereby realizing the determination of the existence and the type of the traffic signal lamp.

The embodiment of the disclosure provides a track data processing method, a track data processing device, track data processing equipment and a computer readable storage medium.

The embodiment of the disclosure provides a track data processing method, which includes: obtaining track data and road network data, wherein the track data comprises vehicle track point data in a preset range of an intersection; extracting road condition features of the intersection from the track data, wherein the road condition features comprise track point density features which indicate a distribution rule of vehicle track points in a direction of advancing to the intersection; obtaining the complexity characteristics of the intersection based on the road network data; and determining whether a motor vehicle signal lamp exists at the intersection or not based on the road condition characteristics and the complexity characteristics of the intersection.

An embodiment of the present disclosure provides a trajectory data processing apparatus, including: the system comprises a data acquisition module, a road network data acquisition module and a road network data acquisition module, wherein the track data comprises vehicle track point data in a preset intersection range; an intersection feature extraction module configured to extract road condition features of the intersection from the trajectory data, the road condition features including track point density features indicating a vehicle track point distribution law in a direction of travel to the intersection; the complexity feature extraction module is configured to obtain complexity features of the intersection based on the road network data; and the signal lamp determining module is configured to determine whether a motor vehicle signal lamp exists at the intersection or not based on the road condition characteristics and the complexity characteristics of the intersection.

An embodiment of the present disclosure provides a trajectory data processing apparatus, including: one or more processors; and one or more memories, wherein the one or more memories have stored therein a computer-executable program that, when executed by the processor, performs the trajectory data processing method as described above.

Embodiments of the present disclosure provide a computer-readable storage medium having stored thereon computer-executable instructions for implementing a trajectory data processing method as described above when executed by a processor.

Embodiments of the present disclosure provide 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, so that the computer device executes the trajectory data processing method according to the embodiment of the present disclosure.

Compared with the traditional method for identifying by using road images, the method provided by the embodiment of the disclosure uses the track data with higher updating frequency to perform data mining, the mining period is shorter, and the change of the signal lamp of the motor vehicle can be identified more timely.

Compared with the existing motor vehicle signal lamp identification method based on track data for data mining, the method provided by the embodiment of the disclosure can further excavate the type of the motor vehicle signal lamp under the condition that the motor vehicle signal lamp is determined to be present or not.

The method provided by the embodiment of the disclosure performs data mining based on road network data and vehicle driving track data, extracts a plurality of representative characteristics related to the motor vehicle signal lamp for training of a machine learning model, and accordingly determines the existence of the motor vehicle signal lamp. In addition, the method provided by the embodiment of the disclosure further determines the type of the motor vehicle signal lamp based on the road network data and the track data in the specific driving direction under the condition that the motor vehicle signal lamp is determined to exist. According to the method, the accuracy of identifying the motor vehicle signal lamp is improved, the manual operation cost for subsequent verification is saved, and the user experience is remarkably improved based on the traffic scheme provided by more accurate motor vehicle signal lamp identification.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings that are required to be used in the description of the embodiments will be briefly described below. It is apparent that the drawings in the following description are only exemplary embodiments of the disclosure, and that other drawings may be derived from those drawings by a person of ordinary skill in the art without inventive effort.

FIG. 1A is a schematic diagram illustrating a scenario for processing a traffic plan request from a user terminal, according to an embodiment of the present disclosure;

FIG. 1B is a schematic diagram illustrating returning to a different traffic scenario in accordance with an embodiment of the present disclosure;

FIG. 1C is a schematic diagram illustrating an example traffic scenario, according to an embodiment of the present disclosure;

FIG. 2 is a flow chart illustrating a trajectory data processing method according to an embodiment of the present disclosure;

FIG. 3A is a graph showing trace point density profiles at an intersection with/without a motor vehicle signal light according to an embodiment of the present disclosure;

fig. 3B is a flowchart illustrating extracting road condition characteristics of an intersection according to an embodiment of the present disclosure;

FIG. 4A is a flow diagram illustrating extracting complexity features of an intersection according to an embodiment of the present disclosure;

FIG. 4B is a schematic diagram illustrating an example topology between intersections according to an embodiment of the present disclosure;

FIG. 5A is a schematic flow chart diagram illustrating a determination of whether a turn signal is present at an intersection according to an embodiment of the present disclosure;

FIG. 5B is a flow diagram illustrating a determination of whether a turn signal is present at an intersection based on at least a portion of trajectory data in accordance with an embodiment of the present disclosure;

FIG. 5C is a diagram illustrating an example roadway and motor vehicle signal light arrangement according to an embodiment of the present disclosure;

FIG. 6 is a schematic flow chart diagram illustrating motor vehicle signal light identification according to an embodiment of the present disclosure;

FIG. 7 is a schematic diagram illustrating a trajectory data processing device according to an embodiment of the present disclosure;

FIG. 8 shows a schematic diagram of a trajectory data processing device according to an embodiment of the present disclosure;

FIG. 9 shows a schematic diagram of an architecture of an exemplary computing device, according to an embodiment of the present disclosure; and

FIG. 10 shows a schematic diagram of a storage medium according to an embodiment of the present disclosure.

Detailed Description

In order to make the objects, technical solutions and advantages of the present disclosure more apparent, example embodiments according to the present disclosure will be described in detail below with reference to the accompanying drawings. It is to be understood that the described embodiments are merely a subset of the embodiments of the present disclosure and not all embodiments of the present disclosure, with the understanding that the present disclosure is not limited to the example embodiments described herein.

In the present specification and the drawings, steps and elements having substantially the same or similar characteristics are denoted by the same or similar reference numerals, and repeated description of the steps and elements will be omitted. Meanwhile, in the description of the present disclosure, the terms "first", "second", and the like are used only for distinguishing the description, and are not to be construed as indicating or implying relative importance or order.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to be limiting of the invention.

For the purpose of describing the present disclosure, concepts related to the present disclosure are introduced below.

The trajectory data processing method of the present disclosure may be based on Artificial Intelligence (AI). Artificial intelligence is a theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use the knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technique of computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can react in a manner similar to human intelligence. For example, with respect to the artificial intelligence-based trajectory data processing method, it is possible to determine the motor vehicle signal lights on the road and the types thereof in a manner similar to a manner in which a human recognizes various traffic signs and signal lights present on the road by the naked eye. The artificial intelligence enables the track data processing method disclosed by the invention to have the functions of accurately extracting the characteristics related to the motor vehicle signal lamp from the track data and road network data of vehicle driving in real time and determining the existence and the type of the motor vehicle signal lamp based on the characteristics by researching the design principle and the implementation method of various intelligent machines.

The trajectory data processing method of the present disclosure may be based on an Intelligent Transportation System (ITS). The Intelligent Transportation System is a comprehensive Transportation System which effectively and comprehensively applies advanced scientific technologies (information technology, computer technology, data communication technology, sensor technology, electronic control technology, automatic control theory, operational research, artificial intelligence and the like) to Transportation, service control and vehicle manufacturing and strengthens the relation among vehicles, roads and users, thereby forming the comprehensive Transportation System which ensures safety, improves efficiency, improves environment and saves energy.

Optionally, the trajectory data processing method of the present disclosure may be based on data mining techniques. The data mining technology utilizes large-scale data generated by daily traffic behaviors to mine implicit patterns in the large-scale data so as to provide powerful and flexible analysis and processing functions for the large-scale data. As an application-oriented data analysis processing technology, the data mining technology can quickly, effectively and deeply analyze mass traffic information and mine traffic patterns implicit in mass traffic data. In the traffic data processing process, the data mining technology summarizes the association among the data by analyzing the potential relationship among the attributes, improves the depth of the data and derives a plurality of relationship data, and then can judge and predict the current state, provide guidance for traffic management and give full play to the value of the data. For example, in the trajectory data processing method disclosed by the invention, by analyzing the trajectory data of vehicle driving and the potential relation between the road network data and the signal lights of the motor vehicles, a plurality of characteristics which can represent the potential relation most are extracted to represent the association between the data and the signal lights of the motor vehicles, so as to fully utilize the potential value of the trajectory data and the road network data.

Optionally, the training model employed by embodiments of the present disclosure may be an artificial intelligence based machine learning model. In general, machine learning models based on artificial intelligence can be classified into supervised learning, unsupervised learning and reinforcement learning based on a learning form, wherein supervised learning is learning a model from labeled training data to make predictions about unknown or future data, and the term "supervised" refers to output signals or labels required by known samples. The present disclosure may employ a supervised machine learning model, train the model based on labeled (present/absent) samples, and then use the trained model to predict the likelihood of motor vehicle signal lights at a new intersection. The supervised machine learning model may include algorithm models such as eXtreme Gradient Boosting (XGBoost), the XGBoost is a Tree model algorithm derived through ensemble learning (ensemble learning), and may be regarded as integration of different single Decision trees (Decision trees, DTs), input features are converted through enhanced DTs, results of all trees are spliced to obtain new combined features, which is proved to be capable of realizing very strong feature conversion. Thus, embodiments of the disclosure may optionally be trained hereinafter using the above-described XGBoost model, but it should be understood that other machine learning models that may achieve similar technical effects are equally applicable to the methods of the disclosure, and that the XGBoost model is used in the disclosure as an example only and not a limitation.

In summary, the solutions provided by the embodiments of the present disclosure relate to artificial intelligence, data mining, and the like, and the embodiments of the present disclosure will be further described with reference to the accompanying drawings.

Fig. 1A is a schematic diagram illustrating a scenario 100 of processing of a traffic plan request from a user terminal according to an embodiment of the present disclosure. Fig. 1B is a schematic diagram illustrating returning to a different traffic scenario in accordance with an embodiment of the present disclosure. Fig. 1C is a schematic diagram illustrating an example traffic scenario, according to an embodiment of the present disclosure.

Currently, there are applications (e.g., various applications typified by an Tencent map, etc.) that generate a plurality of traffic plans for a start point and an end point input by a user, and the user can initiate a traffic plan request among these applications installed on his user terminal. The user terminal may then transmit its input data, such as location information of the start and end points, etc., to the server of the application via the network.

Optionally, the user terminal may specifically include a smartphone, a tablet, a laptop portable computer, a vehicle-mounted terminal, a wearable device, and so on. The user terminal may also be a client that installs a browser or various applications, including system applications and third party applications. The network may be an Internet of Things (Internet of Things) based on the Internet and/or a telecommunication network, which may be a wired network or a wireless network, for example, which may be a Local Area Network (LAN), a Metropolitan Area Network (MAN), a Wide Area Network (WAN), a cellular data communication network, or other electronic networks capable of implementing information exchange functions. The server may be 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 cloud service, a cloud database, cloud computing, a cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, CDN, and a big data and artificial intelligence platform.

As shown in fig. 1A, the server may determine a possible plurality of traffic plans online based on the received input data, and determine an optimal traffic plan therefrom as a recommended plan for the user, so that the user may easily obtain the optimal plan while freely selecting other traffic plans as needed. Subsequently, the server may return the determined plurality of transportation schemes to the user terminal through the network to be displayed to the user.

The interfaces displaying the plurality of transportation schemes to the user may be interfaces (a) and (B) as shown in fig. 1B. The user may request a transportation plan by entering a start point and an end point location in the upper address boxes in interface (a) or (B), respectively (e.g., in fig. 1B, the start point is the user's current location (shown as "my location") and the end point is the a subway station). The server may determine different transportation schemes for the user according to the selection of the travel mode by the user, for example, the transportation scheme when the user selects the walking mode may include a plurality of paths which cannot be passed by a motor vehicle. Since the purpose of the present disclosure is to determine the presence and type of motor vehicle signal lights, the following is described with respect to a driving mode.

In interfaces (a) and (b), three traffic plans (traffic plan one, traffic plan two, and traffic plan three) are shown returning from the server, where each traffic plan includes a corresponding travel time, travel distance, and number of motor vehicle signal lights traveled. The interfaces (a) and (b) respectively illustrate respective driving routes by taking a first traffic scheme and a third traffic scheme as examples, wherein the current position of the user, namely the starting point, is represented by a positioning symbol, the destination position, namely the terminal point, is represented by a mark 'terminal', and a black bold curve between the starting point and the terminal point represents the driving route corresponding to the current traffic scheme.

As shown in interfaces (a) and (b), the traffic plan one is an optimal traffic plan (i.e., recommended plan) with a travel time of 23 minutes and a travel distance of 12 km, and the number of signal lights of passing vehicles is 11. Traffic scenario three has a shorter travel distance (11.5 km) but more motor vehicle signal lights (15) than traffic scenario one, and its predicted travel time is 24 minutes, which may be because the travel distances of traffic scenario one and traffic scenario three are not very different, but traffic scenario three needs to pass more motor vehicle signal lights, resulting in an increase in the travel time required. Traffic scenario two has the same distance traveled (12 km) as traffic scenario one but more motor vehicle signal lights (20) than traffic scenario three, which may travel longer (26 minutes).

As described above, the number of motor vehicle signal lights on the driving route of the vehicle may directly affect the driving time of the vehicle and the selection of the traffic plan, so that accurate identification of the motor vehicle signal lights on the road is of great significance for the selection of the traffic plan and the user experience.

In the existing motor vehicle signal lamp identification technology, a road image is generally used for identification, the technology takes the road image as an input, but because the available road image is generally updated for a long time (generally, the available road image is updated for several days, and the available road image is updated for several months), the motor vehicle signal lamp in practice may be adjusted frequently according to city planning, the input road image may not be the same as the road condition when a user initiates a traffic scheme request, and therefore, the motor vehicle signal lamp identification technology based on the road image may provide a traffic scheme with errors in the driving process of a vehicle, and may even cause danger. In addition, since the identification technology identifies the motor vehicle signal lamp as a point based on the road image, the specific road direction on which the motor vehicle signal lamp acts cannot be judged. For example, as shown in fig. 1C, for an intersection connecting A, B, C, D and E five roads, a vehicle traveling on the road E in the direction of the arrow thereon is not controlled by the traffic light of the intersection, but the identification method based on the road image can only identify whether there is a traffic light at the intersection, and cannot identify which road the traffic light specifically controls.

Therefore, a technology for mining vehicle driving track data on a road to optimize motor vehicle signal lamp identification is also provided, but the two motor vehicle signal lamp identification methods can only determine whether motor vehicle signal lamps exist at an intersection and cannot judge the specific driving direction acted by the motor vehicle signal lamps, namely cannot determine the types of the motor vehicle signal lamps. However, during actual vehicle driving, the determination of the type of motor vehicle signal lights is very necessary because although the right turn of the vehicle at most of the intersection is not controlled by the motor vehicle signal lights, if at the intersection with signal lights controlling the right turn, the above-mentioned identification method may have errors in providing the path planning and navigation to the user, resulting in the driver not actually driving according to the truly optimal traffic scheme, and may also result in the driver violating the traffic regulations or even causing danger by ignoring the motor vehicle signal lights controlling a specific driving direction.

The present disclosure is based on the foregoing, and provides a trajectory data processing method, which performs data mining on both trajectory data and road network data of vehicle driving, extracts various representative features for machine learning model training, and thereby realizes determination of the existence of a traffic signal and the type thereof.

Fig. 2 is a flow chart illustrating a trajectory data processing method 200 according to an embodiment of the present disclosure.

In step 201, trajectory data, which may include vehicle trajectory point data within a predetermined range of an intersection, and road network data may be obtained.

According to the embodiments of the present disclosure, traffic data can be classified into two major categories, static traffic data and dynamic traffic data. The static traffic data may be data that is not updated or is updated at a slow speed after the setting is completed, that is, the road network data in the present disclosure may mainly include, for example, basic geographic information (e.g., urban road network, intersection layout, etc.), management entity information (e.g., electronic eye arrangement, speed limit sign, pedestrian crossing line setting, etc.), traffic management information (area boundary, safety range, traffic police distribution, post, law enforcement station, etc.), traffic passenger information (e.g., passenger line, bus stop information, line radiation diagram, traffic junction, etc.), basic spatial data of urban traffic, and road basic information (e.g., road grade, length, toll information, etc.), etc. In determining whether there is a motor vehicle signal at an intersection, the road network data may be used as a reference, for example, at an intersection where urban roads are more complex (for example, an intersection where A, B, C, D and E five roads are shown in fig. 1C), or an intersection where traffic management is more strict (for example, an intersection where more management entities are provided or an intersection where more management entities are provided are connected), it may be more inclined to provide a motor vehicle signal. The complexity of the intersection will be described below by taking the aforementioned two points (basic geographic information and management entity information) as an example of the road network data to be used, not by way of limitation.

According to the embodiment of the disclosure, the dynamic traffic data may include data acquired by remote sensing means such as satellite remote sensing, aerial photogrammetry, a low-altitude unmanned emergency platform, a ground survey vehicle, a ground video and the like, and traffic data acquired by sensing equipment such as a video, a mobile phone, a bus card, a ground induction coil and the like and a mobile terminal in a ground intelligent traffic system. The track data used by the motor vehicle signal lamp identification method disclosed by the invention can be acquired from the acquired traffic data, and comprise vehicle track data, vehicle real-time position data, vehicle track point data in a preset intersection range and the like.

The driving track of the vehicle is composed of a continuous sequence of track points, and the track points are points collected by positioning the vehicle during driving (it is assumed here that the collection of the track points of the vehicle driving has a certain periodicity, such as once every second or once every five seconds), and include information of longitude, latitude, time, etc. of the positioning. Therefore, after the collected track points are matched with the road network, the track points can be known on which road. When a vehicle runs on a road, if the vehicle is controlled by a signal lamp on a certain intersection with the signal lamp of the motor vehicle, the vehicle needs to stop when a red light appears, and the vehicle can pass under the condition of a green light, track points on the running track of the vehicle are generally in non-uniform distribution. Because at the intersection where the motor vehicle signal lamp exists, when the vehicle waits for the green light, the number of track points in the waiting area can be gradually increased along with the time, and after the red light is changed into the green light, the number of track points in the waiting area can be gradually reduced. Therefore, at the intersection where the motor vehicle signal lamp exists, the density distribution of the track points may have certain periodicity along with the change of the motor vehicle signal lamp, and at the intersection where the motor vehicle signal lamp does not exist, the distribution of the track points is generally uniform, and no obvious periodicity exists.

Alternatively, the intersection predetermined range may be a road area within a predetermined distance range from the intersection on the road in the direction of travel to the intersection, and the predetermined distance range may be selected according to the periodicity of the track point density distribution, so as to better present the relationship between the track point density and the distance from the intersection on the road in the direction of travel to the intersection, that is, the periodicity of the vehicle track point density distribution within the intersection predetermined range.

Fig. 3A is a graph showing a trace point density distribution at an intersection with/without a motor vehicle signal light according to an embodiment of the present disclosure.

Alternatively, for an intersection with a motor vehicle signal lamp, the vehicle waits in a parking area (i.e. a waiting area) during a red light period, and for an intersection without a motor vehicle signal lamp, the vehicle can directly pass through the intersection without waiting, so that whether a motor vehicle signal lamp exists at the intersection can be judged by comparing the density distribution of track points in the waiting area by using the difference.

Accordingly, in step 202, road condition characteristics of the intersection can be extracted from the trajectory data, and the road condition characteristics include track point density characteristics indicating a regular distribution of vehicle track points in a direction of traveling toward the intersection.

As shown in fig. 3A, at an intersection having a motor vehicle signal light, a large number of track points may be gathered on a road in a direction of traveling toward the intersection due to the influence of the motor vehicle signal light, and the track points may generally exhibit a periodicity of peaks and valleys, which is generally related to a period of variation of the motor vehicle signal light. For example, in fig. 3A (a), distinct peaks appear at intersections 0m, 30m, and 50m from the motor vehicle signal light, respectively, and the peaks may appear due to the switching of the motor vehicle signal light, for example, the motor vehicle signal light switches from a red light to a green light at the peaks, so that the speed of the vehicle originally in a stopped or slowly traveling state gradually increases, and the track point density decreases accordingly.

The appearance of a trough after a wave crest may be due to the switching of the motor vehicle signal lights from green to red, resulting in a gradual reduction in the speed of the vehicle behind the trough, with a consequent increase in the track point density, but since the degree of blockage in places further away from the crossing is generally lower, the increased track point density value behind the trough is generally less than the track point density value of the preceding wave crest. Thus, the distance between two peaks can be understood as the distance traveled by the vehicle in one turn of the signal light of the motor vehicle (including first switching from red to green and then from green to red), which can reflect the degree of traffic smoothness at the intersection.

As shown in fig. 3A (b), at an intersection without a signal lamp of a motor vehicle, although the track point density decreases with the distance from the intersection, the track points are distributed more uniformly and have smaller density values, and no obvious wave peak or wave trough occurs.

Thus, an intersection with a motor vehicle signal light can be determined on the basis of the above-mentioned vehicle track point distribution law in the direction of travel to the intersection, which can be indicated by the track point density feature. Optionally, the trace point density feature may be determined based on statistical information of vehicle trace point densities within a predetermined range of the intersection in the trace data.

Optionally, the trace point density features extracted by the embodiment of the present disclosure may include, for example, at least one of a ratio of adjacent peak to corresponding trace point densities at a trough, a ratio of absolute distances between adjacent peaks, a ratio of trace point densities of adjacent peaks, and the like in the above trace point density distribution. It should be understood that, in addition to the above-mentioned trace point density features, other trace point density features that can reflect the trace point density distribution law at the intersection of the motor vehicle signal lamp can also be adopted in the trace data processing method of the present disclosure, and the above-mentioned several trace point density features are used as examples only and are not limited.

Fig. 3B is a flowchart illustrating extracting road condition characteristics of an intersection according to an embodiment of the present disclosure.

According to an embodiment of the present disclosure, step 202 may include the steps as shown in fig. 3B.

In step 2021, a distribution of trajectory points in a direction of travel to the intersection may be determined based on the trajectory data.

As described above, the trajectory data may be used to determine the trajectory point distribution within the predetermined range of the intersection as shown in fig. 3A, and the trajectory point distribution may be used as a basis for extracting important features of subsequent portions and may reflect information related to characteristics of the intersection having the motor vehicle signal lamp.

According to an embodiment of the present disclosure, the trajectory data may be vehicle trajectory point data acquired at each sampling time point within a predetermined time range.

According to an embodiment of the present disclosure, step 2021 may comprise: adding the vehicle track point data obtained by each sampling time point in a preset time range to obtain a track point data accumulation result in the preset time range; extracting a track point data accumulation result in the direction of advancing to the intersection based on the track point data accumulation result; and determining the track point distribution based on the track point data accumulation result in the advancing direction of the intersection.

Alternatively, the number of track points at each position in the direction of traveling to the intersection within the predetermined time range may be counted based on the accumulated result of the track point data within the predetermined time range. Adding the vehicle track point data obtained from each sampling time point within the predetermined time range may refer to superimposing the vehicle track point data obtained within the predetermined time range on the same coordinate plane, so as to obtain a vehicle track point data accumulation result within the predetermined time range, such as the track point density distribution map shown in fig. 3A.

In step 2022, the track point density characteristics and the vehicle kinematics characteristics within the predetermined range of the intersection can be determined from the track point distribution, and the vehicle kinematics characteristics indicate the movement state of the vehicle through the intersection.

Optionally, the trace point density features may be determined based on statistical information of the trace point distribution as described above with reference to fig. 3A.

Alternatively, the vehicle kinematic characteristics within the predetermined range of the intersection may include kinematic information of the vehicle within the predetermined range of the intersection, for example, the vehicle kinematic characteristics may include at least one of statistical information of an average time of the vehicle passing through the intersection (such as a mean and a quantile) and statistical information of an average speed of the vehicle passing through the intersection (such as a mean and a quantile), where the average time of the vehicle passing through the intersection/the average speed may be an average time of the vehicle within the predetermined range of the intersection/the average speed. It should be understood that other vehicle kinematic features capable of reflecting the moving state of the vehicle passing through the intersection may be employed in the trajectory data processing method of the present disclosure as well, in addition to the above-described vehicle kinematic features, which are merely used as examples and are not limiting.

In step 2023, a basic attribute feature of the intersection may be determined according to the track point distribution and the road network data, where the basic attribute feature indicates a relationship between a lane at the intersection and the number of tracks passing through the intersection.

According to an embodiment of the present disclosure, step 2023 may comprise: determining the number of lanes at the intersection based on the road network data; determining the number of tracks at the intersection and the number of tracks on each lane at the intersection based on the distribution of the track points and the number of lanes at the intersection; and determining the basic attribute characteristics of the intersection based on the number of the tracks at the intersection and the number of the tracks on each lane at the intersection.

Optionally, the number of trajectories at the intersection, i.e. the number of vehicles passing through the intersection, which consist of a corresponding plurality of vehicle trajectory points, can be determined on the basis of the trajectory point distribution.

According to an embodiment of the present disclosure, the basic attribute feature may reflect at least one of a degree of vehicle congestion at an intersection, a degree of vehicle congestion at each lane at an intersection, and the like. For example, the degree of vehicle congestion of each lane at the intersection may be a basic use case for each lane at the intersection.

For example, the basic attribute feature may include at least a part of the total number of trajectories through the intersection, the total number of lanes at the intersection, and the average number of trajectories on each lane (i.e., the total number of trajectories through the intersection/the total number of lanes at the intersection), and the like. Therefore, the basic attribute feature of the intersection can reflect the use frequency of each lane at the intersection, and the use frequency is generally associated with the importance degree of the road, and the intersection of the road with higher use frequency is more likely to be provided with the motor vehicle signal lamp, so the basic attribute feature of the intersection can be used as another reference for determining the motor vehicle signal lamp at the intersection.

It should be understood that, besides the above-mentioned basic attribute features, other basic attribute features (for example, traffic flow, etc.) capable of reflecting the use conditions of each lane at the intersection may also be adopted in the trajectory data processing method of the present disclosure, and the above-mentioned several basic attribute features are merely used as examples and are not limited.

According to the embodiment of the disclosure, the road condition characteristics of the intersection may further include the vehicle kinematic characteristics and the basic attribute characteristics.

As described above, the road condition characteristics of the intersection may include track point density characteristics, vehicle kinematics characteristics, and basic attribute characteristics of the intersection, which may reflect the degree of traffic smoothness at the intersection. In addition to these road condition characteristics, other characteristics that may represent traffic characteristics of the intersection having motor vehicle signal lights may also be included in the intersection characteristics of the intersection, which are used as examples only and are not limiting.

In addition, in the track data acquisition, the number of the tracks acquired at some intersections may be insufficient or even no tracks are acquired, so that the acquired road condition characteristics are not significant. In this case, the trajectory data processing method of the present disclosure may predict whether there is a motor vehicle signal lamp at the intersection using the complexity at the intersection.

In step 203, complexity characteristics of the intersection can be obtained based on the road network data.

Fig. 4A is a flowchart illustrating extracting complexity features of an intersection according to an embodiment of the present disclosure.

According to an embodiment of the present disclosure, step 203 may include the steps as shown in fig. 4A.

In step 2031, a topology structure and special attributes of the intersection may be determined based on the road network data, wherein the topology structure includes at least one other intersection connected to the intersection and a road between the intersection, the special attributes are related to a specific physical entity at the intersection, and the specific physical entity includes at least one of a pedestrian crossing, an electronic eye and a speed limit sign.

According to the embodiment of the present disclosure, since intersections generally have a relatively complex relationship structure, which can be connected with surrounding intersections to form a topology of the intersections, the topology of the intersections can be determined based on basic geographic information (e.g., urban road network, intersection layout, etc.) in the road network data described previously. Fig. 4B is a schematic diagram illustrating an example topology between intersections according to an embodiment of the present disclosure.

As shown in (a) and (B) in fig. 4B, where (a) is an actual map showing connection relations of each intersection, taking prediction of a signal light of a motor vehicle at intersection a as an example, and (B) is a road network topology structure centering on intersection a constructed based on the map, intersection a can be connected with intersections B, C and D through roads. Generally, the length of a road between intersections can reflect the complexity of an intersection to some extent, for example, if an intersection is far away from surrounding intersections and the number of tracks passing through the intersection is small (the amount of traffic flow at the intersection is small), the probability that a motor vehicle signal lamp is provided at the intersection is low. Generally, if the relationship characteristics of the intersection are complex, the traffic situation is also complex, so that the motor vehicle signal lamps may be more inclined to be arranged at the intersection in order to maintain the traffic order of the intersection.

Further, various attribute information at an intersection may also reflect the complexity of determining the intersection, such as management entity information in the road network data described previously (e.g., electronic eye arrangement, speed limit sign, pedestrian crosswalk setting, etc.). Because traffic management is generally more stringent at intersections where urban roads are more complex (e.g., intersection a in fig. 4B, etc.), traffic management is often more stringent (e.g., more management entities are located at intersections or on roads connected to intersections), there may be a greater tendency to locate motor vehicle signal lights at these intersections. Therefore, the attribute information of the intersection can be used as a reference for determining the complexity of the intersection.

Therefore, as described above, the trajectory data processing method of the present disclosure extracts a feature capable of representing the complexity of an intersection (i.e., a complexity feature of the intersection) based on the road network data as another important feature for predicting whether there is a motor vehicle signal at the intersection.

In step 2032, the complexity characteristics of the intersection can be obtained through a pre-trained graph embedding model based on the topology and special attributes of the intersection.

Alternatively, the trajectory data processing method of the present disclosure may obtain a graph embedding model capable of outputting multidimensional features representing the complexity of the intersection based on the input graph information, in a graph embedding deep web learning manner. The essence of the method for obtaining the graph embedding model based on the graph embedding deep web learning is to represent the whole graph by one multi-dimensional vector, which is the purpose of the present disclosure to adopt the method (i.e. obtaining the complexity characteristics based on the road network topology and the node attribute information), so it should be understood that other methods that can achieve the same purpose are equally applicable to the present disclosure, and the method for obtaining the graph embedding model based on the graph embedding deep web learning is only used as an example and not a limitation.

According to the embodiment of the present disclosure, taking prediction of the signal lights of the motor vehicles at the intersection a as an example, a road network topology structure with the intersection a as a center as shown in (B) in fig. 4B may be constructed, and B, C, D three intersections are respectively connected to the intersection a through corresponding roads.

According to the embodiment of the present disclosure, A, B, C, D four intersections can be taken as points in the graph network, the length of the road between the intersections is taken as the weight of the edge between the points, and a matrix containing the information of the connection relationship of A, B, C, D four intersections and the length of the road between the intersections is constructed as the input of the graph embedding model employed by the present disclosure. In addition, each point can also have own attribute information, including but not limited to information such as whether a pedestrian crossing line is arranged, whether an electronic eye is arranged, whether a speed limit sign is arranged, and the like, and the attribute information can also be input into the graph embedding model in a matrix form, so that the graph embedding model obtained through training can output more excellent complexity characteristics.

It should be appreciated that in addition to being used in situations where the above-obtained road condition characteristics are not significant, the intersection complexity characteristics can also be used in conjunction with the intersection road condition characteristics for more accurately determining whether a motor vehicle signal is present at the intersection.

In step 204, it can be determined whether there is a signal light of a motor vehicle at the intersection based on the road condition characteristics and the complexity characteristics of the intersection.

According to an embodiment of the present disclosure, the determining whether the motor vehicle signal lamp exists at the intersection based on the road condition characteristic and the complexity characteristic of the intersection in step 204 may include determining whether the motor vehicle signal lamp exists at the intersection through a pre-trained machine learning model based on the road condition characteristic and the complexity characteristic of the intersection.

According to an embodiment of the present disclosure, the machine learning model may be obtained by pre-training based on a sample set of trajectory data and road network data, and the machine learning model takes the road condition features and complexity features extracted from the sample set of trajectory data and road network data as input, so as to determine whether there is a motor vehicle signal lamp as output.

According to an embodiment of the present disclosure, the machine learning model may be an extreme gradient boosting model, wherein the extreme gradient boosting model takes a road condition characteristic and a complexity characteristic of the intersection as input, constructs a new combined characteristic based on the road condition characteristic and the complexity characteristic of the intersection, and outputs a determination of whether there is a motor vehicle signal lamp at the intersection based on the constructed combined characteristic and the road condition characteristic and the complexity characteristic of the intersection.

Optionally, the extracted road condition features and complexity features can be used as input samples of the XGBoost model, and whether motor vehicle signal lamps exist at the intersection or not can be used as sample labels of the input samples and transmitted into the XGBoost model for training.

For example, the XGBoost model may be utilized to discover and construct new valid combined features based on existing original features to determine whether a motor vehicle signal is present at the intersection based on the combined features together with the original features.

When a machine learning model for judging whether motor vehicle signal lamps are arranged at intersections is trained, the proportion of positive and negative samples of a sample set of actually acquired track data and road network data is usually unbalanced, namely the number of samples of the intersections with the motor vehicle signal lamps is far less than that of the samples of the intersections without the motor vehicle signal lamps, but the excessive fitting of samples with large proportion is caused by the unbalanced positive and negative samples, so that the prediction result is biased to the classification with more samples. In addition, an undersampling method can be adopted in model training, so that the difference of the number of positive and negative samples is reduced.

Therefore, the newly obtained road condition characteristics and complexity characteristics of the intersection are input into the trained XGboost model, and the determination of whether the intersection has the motor vehicle signal lamp or not can be output in real time.

As described previously, the motor vehicle signal lamp identification method of the embodiment of the present disclosure may further determine the arrow information of the motor vehicle signal lamp at the intersection (i.e., whether the motor vehicle signal lamp is a direction indication signal lamp controlling a specific driving direction) by using the lane information about the intersection and the periodicity of the distribution of the track points along each driving direction in the road network data, in addition to determining whether the motor vehicle signal lamp exists at the intersection. For example, if there is a right turn lane at an intersection and the track point distribution of the right turn has periodicity, there is a high probability that a direction indication signal lamp for controlling the driving direction of the right turn exists at the intersection.

Therefore, according to the embodiment of the present disclosure, in addition to the step 201 and the step 204, the trajectory data processing method 200 may further include a step 205, that is, in the case that it is determined that the motor vehicle signal exists at the intersection, it may be determined whether a direction indicator signal exists at the intersection based on the road network data and at least a part of the trajectory data.

According to an embodiment of the present disclosure, at least a part of the trajectory data is associated with at least one trajectory at the intersection in a particular trajectory direction, the direction indication signal lamp corresponding to the particular trajectory direction. Wherein the specific track direction may include at least one of a right turn, a left turn, and a u-turn. The specific track direction is the specific driving direction, that is, at the intersection, the right-turn, left-turn or turning around of the vehicle may be performed according to the indication of the traffic light.

Fig. 5A is a schematic flow chart diagram illustrating a determination of whether a turn signal is present at an intersection according to an embodiment of the present disclosure. Fig. 5B is a flow diagram illustrating determining whether a turn signal is present at an intersection based on at least a portion of trajectory data according to an embodiment of the present disclosure. Fig. 5C is a diagram illustrating an example roadway and motor vehicle signal light arrangement according to an embodiment of the present disclosure.

According to an embodiment of the present disclosure, the above determining whether a direction indicator light exists at the intersection based on the road network data and at least a part of the trajectory data may include steps as shown in fig. 5A.

In step 2051, it may be determined whether there is a lane corresponding to the specific track direction at the intersection based on the road network data.

Alternatively, in the case that it is determined that there is a motor vehicle signal lamp at the intersection, it may be determined first whether there is a lane corresponding to a specific track direction at the intersection, and if so, it may be further determined whether there is a direction indication signal lamp for controlling a vehicle traveling in the specific track direction on the lane; conversely, if there is no such lane, it may be considered that the intersection is not provided with the direction indication signal lamp corresponding to such lane.

In step 2052, in a case where it is determined that there is a lane corresponding to the specific track direction, it is determined whether there is a direction indicator light at the intersection based on the track data associated with the at least one track in the track data and the complexity feature of the intersection.

Alternatively, the operation in step 2052 may be understood as an operation similar to the operation in steps 202 and 204, with the difference that the operation in step 2052 indicates a signal light for a direction controlling a specific trajectory direction, thus only the data part of the trajectory data associated with at least one trajectory at the intersection in the specific trajectory direction is used. For example, as shown in fig. 5C, if it is determined whether or not there is a turn signal lamp controlling a right turn driving direction at intersection a, the road condition characteristic associated with the right turn driving direction at intersection a can be obtained by using a data portion related to a right turn trajectory at the intersection in the trajectory data, that is, trajectory data within a predetermined range of the intersection in the direction from intersection a to intersection B.

According to an embodiment of the present disclosure, the above step 2052 may include steps as shown in fig. 5B.

In step 20521, a distribution of vehicle trajectory points associated with the particular trajectory direction in a direction of travel to the intersection may be determined based on trajectory data associated with the at least one trajectory in the trajectory data.

Optionally, the track point density distribution along the specific track direction within the intersection predetermined range may be calculated based on data associated with at least one track along the specific track direction at the intersection in the track data, which reflects a vehicle track point distribution rule along the specific track direction within the intersection predetermined range.

In step 20522, a trajectory point density characteristic and a vehicle kinematics characteristic of the intersection associated with the at least one trajectory may be determined based on the trajectory point distribution, the vehicle kinematics characteristic being indicative of a kinematics characteristic of a vehicle passing through the intersection in the particular trajectory direction.

Optionally, a trace point density characteristic associated with at least one trace for the intersection may be determined based on the statistical information of the trace point distribution as described above with reference to fig. 3A.

Alternatively, the vehicle kinematic characteristics within the predetermined range of the intersection may include motion information of the vehicle in a specific track direction within the predetermined range of the intersection, for example, statistical information of average time of the vehicle passing through the intersection in the specific track direction (such as a mean and a quantile) and statistical information of average speed of the vehicle passing through the intersection in the specific track direction (such as a mean and a quantile).

In step 20523, a basic attribute feature of the intersection may be determined according to the track point distribution and the road network data, where the basic attribute feature indicates a relationship between a lane at the intersection along the specific track direction and the number of tracks passing through the intersection along the specific track direction, and the road condition feature of the intersection further includes the vehicle kinematics feature and the basic attribute feature.

Alternatively, a base attribute feature of the intersection associated with a particular track direction may be determined based on the distribution of track points along the particular track direction and associated road network data, which may reflect a base usage of a lane at the intersection associated with the particular track direction.

For example, the basic attribute features may include the total number of tracks passing through the intersection in a specific track direction, the total number of lanes at the intersection associated with the specific track direction, and the average number of tracks on the lane associated with the specific track direction (i.e., the total number of tracks passing through the intersection/the total number of lanes at the intersection), and so on.

In step 20524, it may be determined whether a turn signal exists at the intersection through a pre-trained machine learning model based on the road condition characteristics and the complexity characteristics of the intersection associated with the specific track direction.

According to an embodiment of the present disclosure, the machine learning model may be an extreme gradient boost model, wherein the extreme gradient boost model takes as input a road condition characteristic and a complexity characteristic of the intersection associated with the specific trajectory direction and takes as output a determination of whether there is a turn signal at the intersection.

Optionally, the extracted road condition features and complexity features associated with a specific track direction may be used as input samples of the XGBoost model, and whether a corresponding direction indication signal lamp exists at an intersection or not may be used as a sample label of the input sample and transmitted into the XGBoost model for training.

Therefore, the road condition characteristics and the complexity characteristics of the newly obtained intersection associated with the specific track direction are input into the trained XGboost model, and the determination of whether the intersection has the corresponding direction indication signal lamp or not can be output in real time. Alternatively, one or more direction indicator lights, each for controlling a different track direction, that may be present at the intersection may be determined, thereby forming a complete identification of the traffic lights present at the intersection.

Fig. 6 is a schematic flow diagram illustrating motor vehicle signal light identification according to an embodiment of the present disclosure.

As shown in fig. 6, the trajectory data processing method according to the embodiment of the present disclosure mainly uses trajectory data and road network data.

For the original trajectory data, since the acquisition of the trajectory data may be affected by external factors such as networks and devices, and human factors, and some abnormal trajectories may exist in the actually acquired trajectory data, the abnormal trajectory data needs to be screened before feature extraction.

Optionally, the operation of screening the abnormal trajectory data may include:

1) too short traces are filtered. In general, if the number of trace points on a trace is less than 3, the trace can be considered to be too short and then can be directly filtered.

2) Tracks that deviate too far from the road are filtered. If a track is too far from the real road network, it may be artificially forged or due to equipment anomalies that collect the track, and so it can be filtered directly.

3) And filtering tracks with more track points which cannot be matched with the road network. If a track has more track points that cannot match the road network, then the track may be artificially forged or caused by equipment anomalies that collect the track, so it can be directly filtered.

Therefore, for the preprocessed trajectory data, feature extraction can be performed to obtain road condition data of the intersection, as shown in fig. 6.

For road network data, data related to a road network topology structure taking an intersection to be determined as a center can be input into a trained graph embedding model in a matrix form so as to extract complexity characteristics of the intersection.

Therefore, the extracted road condition characteristics and complexity characteristics of the intersection are input into the trained XGboost model, so that the determination of whether the intersection has the motor vehicle signal lamp or not can be output in real time, and then the determination of whether the intersection has the direction indication signal lamp or not can be output.

Fig. 7 is a schematic diagram illustrating a trajectory data processing device 700 according to an embodiment of the present disclosure.

The trajectory data processing device 700 may include a data acquisition module 701, an intersection feature extraction module 702, a complexity feature extraction module 703, and a signal light determination module 704.

According to an embodiment of the present disclosure, the data acquisition module 701 may be configured to acquire trajectory data including vehicle trajectory point data within a predetermined range of an intersection and road network data.

Alternatively, the intersection predetermined range may be a road region within a predetermined distance range from the intersection on a road in a direction of traveling to the intersection, and the predetermined distance range may be selected according to a periodicity of the track point density distribution to better represent a relationship between the track point density and a distance from the intersection on the road in the direction of traveling to the intersection, that is, the periodicity of the vehicle track point density distribution within the intersection predetermined range.

According to an embodiment of the present disclosure, the intersection feature extraction module 702 may be configured to extract road condition features of the intersection from the trajectory data, the road condition features including track point density features indicating a distribution rule of vehicle track points in a direction of traveling to the intersection.

Alternatively, the intersection feature extraction module 702 extracting the road condition feature of the intersection from the trajectory data may include the operations as described with reference to fig. 3B.

For example, the trace point density features extracted by the embodiment of the present disclosure may include, for example, a ratio of the densities of corresponding trace points at adjacent peaks and troughs, a ratio of absolute distances between adjacent peaks, a ratio of the densities of trace points at adjacent peaks, and the like in the above-described trace point density distribution.

According to an embodiment of the present disclosure, the complexity feature extraction module 703 may be configured to obtain complexity features of the intersection based on the road network data.

Since intersections generally have a relatively complex relationship structure, and they can be connected with surrounding intersections to form their own topology structure, the topology structure of intersections can be determined based on the basic geographic information (e.g., urban road network, intersection layout, etc.) in the road network data described previously.

Alternatively, the complexity feature extraction module 703 obtaining the complexity feature of the intersection based on the road network data may include the operations as described with reference to fig. 4A.

According to an embodiment of the present disclosure, the signal light determination module 704 may be configured to determine whether there is a motor vehicle signal light at the intersection based on the road condition characteristics and the complexity characteristics of the intersection.

Alternatively, the signal lamp determining module 704 determining whether a signal lamp of a motor vehicle exists at the intersection based on the road condition characteristic and the complexity characteristic of the intersection may include determining whether a signal lamp of a motor vehicle exists at the intersection through a pre-trained extreme gradient boost model based on the road condition characteristic and the complexity characteristic of the intersection.

For example, the extracted road condition features and complexity features can be used as input samples of the XGBoost model, and whether motor vehicle signal lamps exist at the intersection or not can be used as sample labels of the input samples and transmitted into the XGBoost model for training. The newly obtained road condition characteristics and complexity characteristics of the intersection are input into the trained XGboost model, so that the determination of whether the intersection has the motor vehicle signal lamp or not can be output in real time.

According to an embodiment of the present disclosure, in addition to the above modules, the trajectory data processing device 700 may further include a signal lamp type determining module 705, which may be configured to determine whether a direction signal lamp exists at the intersection based on the road network data and at least a part of the trajectory data, in case that it is determined that a motor vehicle signal lamp exists at the intersection.

Optionally, the signal lamp type determining module 705 extracting the road condition characteristics of the intersection from the trajectory data may include operations as described with reference to fig. 5A and 5B.

The signal lamp type determining module 705 can further utilize the lane information about the intersection in the road network data and the periodicity of the distribution of the track points along each driving direction to achieve the purpose of determining the type of the motor vehicle signal lamp at the intersection.

According to still another aspect of the present disclosure, there is also provided a trajectory data processing apparatus. Fig. 8 shows a schematic diagram of a trajectory data processing device 2000 according to an embodiment of the present disclosure.

As shown in FIG. 8, the trajectory data processing device 2000 may include one or more processors 2010, and one or more memories 2020. Wherein the memory 2020 has stored therein computer readable code which, when executed by the one or more processors 2010, may perform a trajectory data processing method as described above.

The processor in the embodiments of the present disclosure may be an integrated circuit chip having signal processing capabilities. The processor may be a general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components. The various methods, steps, and logic blocks disclosed in the embodiments of the present application may be implemented or performed. The general purpose processor may be a microprocessor or the processor may be any conventional processor or the like, which may be of the X86 or ARM architecture.

In general, the various example embodiments of this disclosure may be implemented in hardware or special purpose circuits, software, firmware, logic or any combination thereof. Certain aspects may be implemented in hardware, while other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device. While aspects of embodiments of the disclosure have been illustrated or described as block diagrams, flow charts, or using some other pictorial representation, it is well understood that the blocks, apparatus, systems, techniques or methods described herein may be implemented in, as non-limiting examples, hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.

For example, a method or apparatus in accordance with embodiments of the present disclosure may also be implemented by way of the architecture of computing device 3000 shown in fig. 9. As shown in fig. 9, computing device 3000 may include a bus 3010, one or more CPUs 3020, a Read Only Memory (ROM)3030, a Random Access Memory (RAM)3040, a communication port 3050 to connect to a network, input/output components 3060, a hard disk 3070, and the like. A storage device in the computing device 3000, such as the ROM 3030 or the hard disk 3070, may store various data or files used in the processing and/or communication of the trajectory data processing method provided by the present disclosure and program instructions executed by the CPU. Computing device 3000 can also include user interface 3080. Of course, the architecture shown in FIG. 8 is merely exemplary, and one or more components of the computing device shown in FIG. 9 may be omitted as needed in implementing different devices.

According to yet another aspect of the present disclosure, there is also provided a computer-readable storage medium. Fig. 10 shows a schematic diagram 4000 of a storage medium according to the present disclosure.

As shown in fig. 10, the computer storage media 4020 has stored thereon computer readable instructions 4010. The computer readable instructions 4010, when executed by a processor, can perform a trajectory data processing method according to an embodiment of the present disclosure described with reference to the above figures. The computer readable storage medium in embodiments of the present disclosure may be either volatile memory or nonvolatile memory, or may include both volatile and nonvolatile memory. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be Random Access Memory (RAM), which acts as external cache memory. By way of example and not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM), Synchronous Link Dynamic Random Access Memory (SLDRAM), and direct memory bus random access memory (DRRAM). It should be noted that the memories of the methods described herein are intended to comprise, without being limited to, these and any other suitable types of memory. It should be noted that the memories of the methods described herein are intended to comprise, without being limited to, these and any other suitable types of memory.

Embodiments of the present disclosure also provide 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, so that the computer device executes the trajectory data processing method according to the embodiment of the present disclosure.

It is to be noted that 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 disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises at least one executable instruction 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 and/or flowchart illustration, and combinations of blocks in the block diagrams and/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 exemplary embodiments of the present disclosure described in detail above are merely illustrative, and not restrictive. It will be appreciated by those skilled in the art that various modifications and combinations of these embodiments or features thereof may be made without departing from the principles and spirit of the disclosure, and that such modifications are intended to be within the scope of the disclosure.

Claims

1. A trajectory data processing method, comprising:

obtaining track data and road network data, wherein the track data comprises vehicle track point data in a preset range of an intersection;

extracting road condition features of the intersection from the track data, wherein the road condition features comprise track point density features which indicate a distribution rule of vehicle track points in a direction of advancing to the intersection;

obtaining the complexity characteristics of the intersection based on the road network data; and

and determining whether a motor vehicle signal lamp exists at the intersection or not based on the road condition characteristics and the complexity characteristics of the intersection.

2. The method of claim 1, wherein extracting road condition characteristics of the intersection from the trajectory data comprises:

determining a distribution of trajectory points in a direction of travel to the intersection based on the trajectory data;

determining the density characteristics of the track points and vehicle kinematic characteristics in a preset range of the intersection according to the distribution of the track points, wherein the vehicle kinematic characteristics indicate the motion state of a vehicle passing through the intersection; and

determining basic attribute characteristics of the intersection according to the track point distribution and the road network data, wherein the basic attribute characteristics indicate the relationship between lanes at the intersection and the number of tracks passing through the intersection;

wherein the road condition characteristics of the intersection further comprise the vehicle kinematic characteristics and the basic attribute characteristics.

3. The method as claimed in claim 2, wherein the trajectory data is vehicle trajectory point data acquired at respective sampling time points within a predetermined time range;

wherein said determining a distribution of trajectory points in a direction of travel to the intersection based on the trajectory data comprises:

adding the vehicle track point data obtained by each sampling time point in a preset time range to obtain a track point data accumulation result in the preset time range;

extracting a track point data accumulation result in the direction of advancing to the intersection based on the track point data accumulation result; and

and determining the track point distribution based on the track point data accumulation result in the advancing direction of the intersection.

4. The method of claim 2, wherein the basic attribute feature is used to reflect at least one of a degree of vehicle congestion at the intersection, and a degree of vehicle congestion at each lane at the intersection.

5. The method of claim 2, wherein said determining basic attribute characteristics of said intersection from said distribution of trace points and said road network data comprises:

determining the number of lanes at the intersection based on the road network data;

determining the number of tracks at the intersection and the number of tracks on each lane at the intersection based on the distribution of the track points and the number of lanes at the intersection; and

and determining the basic attribute characteristics of the intersection based on the number of the tracks at the intersection and the number of the tracks on each lane at the intersection.

6. The method of claim 1, wherein obtaining complexity characteristics of said intersection based on said road network data comprises:

determining a topological structure and special attributes of the intersection based on the road network data, wherein the topological structure comprises at least one other intersection connected with the intersection and a road between the intersection and the topological structure, the special attributes are related to specific physical entities at the intersection, and the specific physical entities comprise at least one of pedestrian crossing lines, electronic eyes and speed limit signs; and

and obtaining the complexity characteristics of the intersection through a pre-trained graph embedding model based on the topological structure and the special attributes of the intersection.

7. The method of claim 2 or 6, wherein determining whether a motor vehicle signal light is present at the intersection based on the road condition characteristics and the complexity characteristics of the intersection comprises:

determining whether a motor vehicle signal lamp exists at the intersection or not through a pre-trained machine learning model based on the road condition characteristics and the complexity characteristics of the intersection;

the machine learning model is obtained by pre-training based on a sample set of track data and road network data, and takes the road condition characteristics and the complexity characteristics extracted from the sample set of the track data and the road network data as input so as to determine whether a motor vehicle signal lamp exists as output.

8. The method of claim 7, wherein the machine learning model is an extreme gradient boosting model,

the extreme gradient lifting model takes the road condition characteristics and the complexity characteristics of the intersection as input, constructs a new combined characteristic based on the road condition characteristics and the complexity characteristics of the intersection, and outputs the determination of whether the motor vehicle signal lamp exists at the intersection based on the constructed combined characteristic and the road condition characteristics and the complexity characteristics of the intersection.

9. The method of claim 1, further comprising:

determining whether a direction signal lamp exists at the intersection based on at least one part of the road network data and the track data under the condition that the motor vehicle signal lamp exists at the intersection;

wherein at least a portion of the trajectory data is associated with at least one trajectory at the intersection along a particular trajectory direction, the direction indication signal corresponding to the particular trajectory direction;

wherein the specific track direction includes at least one of a right turn, a left turn, and a u-turn.

10. The method of claim 9, wherein determining whether a turn signal is present at the intersection based on the road network data and at least a portion of the trajectory data comprises:

determining whether a lane corresponding to the specific track direction exists at the intersection based on the road network data; and

in a case where it is determined that there is a lane corresponding to the specific track direction, it is determined whether there is a turn signal at the intersection based on the track data associated with the at least one track among the track data and the complexity characteristic of the intersection.

11. The method of claim 10, wherein determining whether a turn signal is present at the intersection based on the trajectory data associated with the at least one of the trajectories and the complexity characteristics of the intersection comprises:

determining a distribution of vehicle trajectory points associated with the particular trajectory direction in a direction of travel to the intersection based on trajectory data associated with the at least one trajectory in the trajectory data;

determining, from the distribution of trace points, trace point density characteristics and vehicle kinematic characteristics of the intersection associated with the at least one trace, the vehicle kinematic characteristics being indicative of kinematic characteristics of vehicles passing through the intersection in the particular trace direction;

determining basic attribute characteristics of the intersection according to the track point distribution and the road network data, wherein the basic attribute characteristics indicate the relationship between the lanes of the intersection along the specific track direction and the track quantity passing through the intersection along the specific track direction, and the road condition characteristics of the intersection further comprise the vehicle kinematics characteristics and the basic attribute characteristics; and

and determining whether a direction indication signal lamp exists at the intersection or not through a pre-trained machine learning model based on the road condition characteristics and the complexity characteristics of the intersection associated with the specific track direction.

12. The method of claim 11, wherein the machine learning model is an extreme gradient boosting model,

and the extreme gradient lifting model takes the road condition characteristic and the complexity characteristic of the intersection, which are associated with the specific track direction, as input, and takes the determination of whether a direction indication signal lamp exists at the intersection as output.

13. A trajectory data processing device comprising:

the system comprises a data acquisition module, a road network data acquisition module and a road network data acquisition module, wherein the track data comprises vehicle track point data in a preset intersection range;

an intersection feature extraction module configured to extract road condition features of the intersection from the trajectory data, the road condition features including track point density features indicating a vehicle track point distribution law in a direction of travel to the intersection;

the complexity feature extraction module is configured to obtain complexity features of the intersection based on the road network data; and

and the signal lamp determining module is configured to determine whether a motor vehicle signal lamp exists at the intersection or not based on the road condition characteristics and the complexity characteristics of the intersection.

14. A trajectory data processing device comprising:

one or more processors; and

one or more memories having stored therein a computer-executable program that, when executed by the processor, performs the method of any of claims 1-12.

15. A computer program product comprising computer instructions which, when executed by a processor, cause a computer device to perform the method of any one of claims 1-12.

16. A computer-readable storage medium having stored thereon computer-executable instructions for implementing the method of any one of claims 1-12 when executed by a processor.