Point of interest processing method and apparatus, electronic device, and computer-readable storage medium

Abstract

The application provides a point of interest processing method and device, electronic equipment and computer readable storage medium, related embodiments can be applied to cloud technology, artificial intelligence, intelligent transportation and various scenes; The method comprises the following steps: determining the parking heat characteristics of a plurality of road segments according to the speed of a plurality of historical driving trajectories; filtering out the passing driving trajectory passing through the point of interest according to the distance between the plurality of historical driving trajectories and the point of interest, to determine the passing heat characteristics of the point of interest; filtering out the passing driving trajectory passing through the point of interest according to the similarity of the orientation angle between the plurality of passing driving trajectories and the point of interest, to determine the passing heat characteristics of the point of interest; performing state prediction processing according to the parking heat characteristics, the passing heat characteristics and the passing heat characteristics, to obtain the state for representing whether the point of interest can pass. Through the application, the implementation cost of determining the state of the point of interest can be reduced, the processing efficiency is improved, and the accuracy of the determined state is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the Internet of Vehicles technology, and in particular to a point of interest processing method and device, an electronic device, and a computer readable storage medium. BACKGROUND

[0002] A point of interest (POI) is a term in a geographic information system, which generally refers to all geographic objects that can be abstracted as points, such as schools, banks, restaurants, gas stations, hospitals, supermarkets, etc. The state for representing whether it is passable is an important attribute of a point of interest, and plays an important role in various applications (such as navigation) of a geographic information system.

[0003] In the scheme provided by the related art, an image of a location where a point of interest is located is usually shot in the field, and key information (such as a “no passing” icon) in the image is extracted through image recognition, and then the state of the point of interest is judged. However, this scheme involves a field collection link, has a high implementation cost, and has a low processing efficiency. SUMMARY

[0004] Embodiments of the present application provide a point of interest processing method, device, electronic device, computer readable storage medium, and computer program product, which can reduce the implementation cost and improve the processing efficiency.

[0005] The technical scheme of the embodiments of the present application is implemented as follows:

[0006] The embodiments of the present application provide a point of interest processing method, which comprises:

[0007] determining a parking heat feature of a road segment respectively passed through by a plurality of historical driving trajectories according to speeds of the plurality of historical driving trajectories;

[0008] screening a passing driving trajectory passing through the point of interest from the plurality of historical driving trajectories according to distances between the plurality of historical driving trajectories and the point of interest, and determining a passing heat feature of the point of interest according to a number of the passing driving trajectories;

[0009] screening a passing driving trajectory passing through the point of interest from a plurality of the passing driving trajectories according to a similarity of an orientation angle between the plurality of the passing driving trajectories and the point of interest, and determining a passing heat feature of the point of interest according to a number of the passing driving trajectories;

[0010] performing state prediction processing according to the parking heat feature, the passing heat feature, and the passing heat feature to obtain a state for representing whether the point of interest is passable.

[0011] The embodiments of the present application provide a point of interest processing device, which comprises:

[0012] a feature mining module, configured to:

[0013] determine, according to speeds of a plurality of historical driving trajectories, a parking heat feature of road segments respectively passed through by the plurality of historical driving trajectories;

[0014] screen, according to distances between the plurality of historical driving trajectories and a point of interest, passing driving trajectories passing through the point of interest from the plurality of historical driving trajectories, and determine, according to a number of the passing driving trajectories, a passing heat feature of the point of interest;

[0015] screen, according to similarity of orientation angles between a plurality of the passing driving trajectories and the point of interest, through driving trajectories passing through the point of interest from the plurality of the passing driving trajectories, and determine, according to a number of the through driving trajectories, a through heat feature of the point of interest;

[0016] a state prediction module, configured to perform state prediction processing according to the parking heat feature, the passing heat feature and the through heat feature, to obtain a state representing whether the point of interest is passable.

[0017] Embodiments of the present application provide an electronic device, comprising:

[0018] a memory, configured to store executable instructions;

[0019] a processor, configured to execute the executable instructions stored in the memory, to implement the point of interest processing method provided by the embodiments of the present application.

[0020] Embodiments of the present application provide a computer readable storage medium, storing executable instructions, configured to cause a processor to execute the point of interest processing method provided by the embodiments of the present application.

[0021] Embodiments of the present application provide a computer program product, comprising executable instructions, configured to cause a processor to execute the point of interest processing method provided by the embodiments of the present application.

[0022] Embodiments of the present application have the following beneficial effects:

[0023] Based on the three aspects of feature mining of the plurality of historical driving trajectories, in the first aspect, the parking heat features of each road segment are mined according to the speed of the historical driving trajectories; in the second aspect, the passing heat features of the interest point are mined according to the distance between the historical driving trajectories and the interest point; and in the third aspect, the penetrating heat features of the interest point are mined according to the similarity of the orientation angle between the passing driving trajectory and the interest point. Finally, the state of the interest point is obtained by combining the three features mined in the state prediction processing. Since the historical driving trajectories are easy to obtain, the implementation cost of determining the state of the interest point is reduced, and the processing efficiency is improved; at the same time, since the three types of features mined have strong relevance with the state of the interest point, the accuracy of the state determined can also be ensured. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 is an architecture schematic diagram of the interest point processing system provided by the embodiments of the present application;

[0025] Figure 2 is an architecture schematic diagram of the terminal device provided by the embodiments of the present application;

[0026] Figure 3A is a flow schematic diagram of the interest point processing method provided by the embodiments of the present application;

[0027] Figure 3B is a flow schematic diagram of the interest point processing method provided by the embodiments of the present application;

[0028] Figure 3C is a flow schematic diagram of the interest point processing method provided by the embodiments of the present application;

[0029] Figure 3D is a flow schematic diagram of the interest point processing method provided by the embodiments of the present application;

[0030] Figure 3E is a flow schematic diagram of the interest point processing method provided by the embodiments of the present application;

[0031] Figure 4A is a flow schematic diagram of the feature mining and state prediction processing provided by the embodiments of the present application;

[0032] Figure 4B is a flow schematic diagram of the feature mining and state prediction processing provided by the embodiments of the present application;

[0033] Figure 5 is a flow schematic diagram of the state determination through a plurality of stages provided by the embodiments of the present application. DETAILED DESCRIPTION

[0034] In order to make the purposes, technical solutions and advantages of the present application clearer, the following further describes the present application in conjunction with the accompanying drawings, the described embodiments should not be regarded as limitations on the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0035] In the following description, reference is made to "some embodiments", which describes a subset of all possible embodiments, but it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict. In the following description, the term "a plurality of" refers to at least two.

[0036] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.

[0037] The relevant data collection and processing in the embodiments of the present application should strictly comply with the requirements of relevant laws and regulations, obtain the informed consent or separate consent of the personal information subject, and within the scope of authorization of laws and regulations and the personal information subject, carry out subsequent data use and processing.

[0038] Before further detailing the embodiments of the present application, the terms and terms involved in the embodiments of the present application are explained, the terms and terms involved in the embodiments of the present application are applicable to the following explanations.

[0039] 1) Interest point: a term in geographic information system, generally refers to all geographic objects that can be abstracted as points, in the embodiments of the present application, can refer to interest points with states for representing whether passable, such as door type interest points (such as the east gate of a certain university, the south gate of a certain factory, the east gate of a certain residential community, etc.), entrance type interest points (such as the exit / entrance of a certain mansion parking lot), etc.

[0040] 2) Historical driving trajectory: refers to the historical driving trajectory of a vehicle, the historical driving trajectory can be calculated according to a series of historical trajectory points. Wherein, the recording method of the trajectory point is not limited, for example, it can be recorded by a device with coordinate recording function in the vehicle (such as a fixed terminal device such as a navigation car machine, or a mobile terminal device installed with a recording positioning function APP).

[0041] 3) Parent interest point: refers to a parent point in logical and geographical relationship, for example, the parent interest point of the interest point "xx university east gate" is "xx university", and the child interest point of "xx university" includes "xx university east gate".

[0042] 4) Artificial Intelligence (AI): is to use digital computers or digital computer controlled machine simulation, extension and expansion of human intelligence, perception of the environment, knowledge acquisition and use of knowledge to obtain the best results of theory, method, technology and application system. Machine Learning (ML) is the core of artificial intelligence, and is the fundamental way to make computers intelligent, and its application covers various fields of artificial intelligence. In the embodiments of the present application, a state prediction model can be constructed based on the principle of machine learning to realize the prediction of the state of the interest point.

[0043] 5) Intelligent Traffic System (ITS): also known as Intelligent Transportation System, is to effectively integrate advanced science and technology (information technology, computer technology, data communication technology, sensor technology, electronic control technology, automatic control theory, operations research, artificial intelligence, etc.) into transportation, service control and vehicle manufacturing, to strengthen the connection between vehicles, roads and users, so as to form a comprehensive transportation system that ensures safety, improves efficiency, improves the environment and saves energy. The embodiments of the present application can be applied to the intelligent traffic system to realize intelligent traffic, for example, the states of various interest points in the electronic map can be updated in time and effectively to support search, navigation and other applications.

[0044] For the state recognition of the interest point, the related technologies mainly provide the following two schemes:

[0045] 1) Artificially collecting and manually making the state of the interest point, for example, relevant personnel go to the location of the interest point to observe and manually update the state of the interest point;

[0046] 2) Taking an image of the location of the interest point and extracting key information (such as "no entry" icon) to determine the state through image recognition.

[0047] The above two schemes both include a field collection link, resulting in high cost, long cycle and low efficiency. According to the investigation, from the cost point of view, the implementation cost of the above two schemes is not less than 5 yuan per interest point; from the efficiency point of view, the production cycle of the above two schemes is about 1 week per interest point, that is, the state of the interest point can be updated only once a week.

[0048] The embodiment of the present application provides a point of interest processing method and device, electronic equipment, computer readable storage medium and computer program product, which can reduce the implementation cost of determining the state of the point of interest, improve the processing efficiency, and also ensure the accuracy of the determined state. The following describes an exemplary application of the electronic equipment provided by the embodiment of the present application. The electronic equipment provided by the embodiment of the present application can be implemented as various types of terminal equipment, or as a server.

[0049] Referring to Figure 1 , Figure 1 is an architecture schematic diagram of a point of interest processing system 100 provided by the embodiment of the present application. The terminal equipment 400 is connected to the server 200 through the network 300. The network 300 can be a wide area network or a local area network, or a combination of the two. The embodiment of the present application can determine the state of the point of interest through the point of interest processing system. The determined state can be applied to various application scenarios. For example, when planning a route, the point of interest with a state of being unpassable (i.e., closed) can be avoided to ensure that the planned route can be driven (or walked); for another example, when receiving a search request, the state of the point of interest can be responded to according to the state of the point of interest to facilitate the user to know the actual state of the point of interest; for another example, in the navigation process, the state change of the point of interest that is passed through or needs to be reached is reminded to facilitate switching the point of interest or re-planning the route to meet the navigation needs of the user.

[0050] In some embodiments, taking the electronic equipment as an example, the point of interest processing method provided by the embodiment of the present application can be implemented by the terminal equipment. For example, the terminal equipment 400 determines a parking heat feature of a road segment respectively passed through by a plurality of historical driving trajectories according to the speed of the plurality of historical driving trajectories; filters out a passing driving trajectory passing through the point of interest from the plurality of historical driving trajectories according to the distance between the plurality of historical driving trajectories and the point of interest, and determines a passing heat feature of the point of interest according to the number of the passing driving trajectories; filters out a passing-through driving trajectory passing through the point of interest from the plurality of passing driving trajectories according to the orientation angle similarity between the plurality of passing driving trajectories and the point of interest, and determines a passing-through heat feature of the point of interest according to the number of the passing-through driving trajectory; and performs state prediction processing according to the parking heat feature, the passing heat feature and the passing-through heat feature to obtain a state for representing whether the point of interest is passable.

[0051] The historical driving trajectory can be pre-stored in the local terminal equipment 400, or obtained by the terminal equipment 400 from the outside world (such as the server 200, a block chain, etc.). The terminal equipment 400 can store the predicted state of the point of interest in the local terminal equipment 400 to facilitate responding to possible search requests, route planning requests (or navigation requests), etc.

[0052] In some embodiments, the electronic device is a server, and the interest point processing method provided in the embodiments of the present application can also be implemented by the server. For example, the server 200 can mine the parking heat feature, the passing heat feature and the crossing heat feature according to a plurality of historical driving trajectories, perform state prediction processing according to the parking heat feature, the passing heat feature and the crossing heat feature, and obtain a state for representing whether the interest point is crossable.

[0053] In some embodiments, the interest point processing method provided in the embodiments of the present application can also be implemented by the terminal device and the server in cooperation. For example, the terminal device 400 (there can be a plurality of terminal devices, and it is not limited to the terminal device 400) can send a historical driving trajectory (or a series of trajectory points) to the server 200, so that the server 200 predicts the state of the interest point according to the historical driving trajectory. The server 200 can store the predicted state of the interest point, such as storing in a distributed file system, a database or a block chain of the server 200 itself, and respond to a search request or a route planning request of the terminal device 400 according to the stored state; or the server 200 can also send the predicted state to the terminal device 400, so that the terminal device 400 responds to the search request or the route planning request according to the locally stored state.

[0054] In some embodiments, various results involved in the interest point processing process (such as trajectory points, historical driving trajectories, various features, states of interest points, etc.) can be stored in a block chain. Since the block chain has the characteristics of being tamper-proof, the accuracy of the data in the block chain can be guaranteed. The electronic device can send a query request to the block chain to query the data stored in the block chain.

[0055] In some embodiments, the terminal device 400 or the server 200 can implement the interest point processing method provided in the embodiments of the present application by running a computer program. For example, the computer program can be a native program or a software module in the operating system; it can be a native application program (APP), that is, a program that needs to be installed in the operating system to run, such as an electronic map APP; it can also be a small program, that is, a program that only needs to be downloaded into a browser environment to run; it can also be a small program that can be embedded into any APP. In summary, the above computer program can be any form of application program, module or plug-in.

[0056] In some embodiments, the server 200 can be a standalone physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and basic cloud computing services such as big data and artificial intelligence platforms. The terminal device 400 can be a smartphone, a tablet computer, a notebook computer, a desktop computer, a smart speaker, a smart watch, a vehicle-mounted terminal, a smart television, and the like, but is not limited thereto. The terminal device and the server can be connected directly or indirectly through wired or wireless communication, and the present application embodiments are not limited thereto.

[0057] For the case where the electronic device is a server, it can be understood that the structure shown in Figure 2 FIG. 1 can be omitted by default. See Figure 2 , Figure 2 FIG. 1 is a structural schematic diagram of a terminal device 400 provided by an embodiment of the present application, Figure 2 The terminal device 400 shown in FIG. 1 includes at least one processor 410, a memory 450, at least one network interface 420, and a user interface 430. The various components in the terminal device 400 are coupled together through a bus system 440. It can be understood that the bus system 440 is used to realize the connection and communication between the components. In addition to including a data bus, the bus system 440 also includes a power bus, a control bus, and a status signal bus. However, for the purpose of clear illustration, all kinds of buses are marked as the bus system 440 in Figure 2 FIG. 1.

[0058] The processor 410 can be an integrated circuit chip with signal processing capability, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0059] The user interface 430 includes one or more output devices 431 that enable the presentation of media content, including one or more speakers and / or one or more visual display screens. The user interface 430 also includes one or more input devices 432, including user interface components that facilitate user input, such as a keyboard, a mouse, a microphone, a touch screen display, a camera, other input buttons and controls.

[0060] The memory 450 can be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid-state memory, hard drives, optical drives, and the like. The memory 450 optionally includes one or more storage devices remotely located from the processor(s) 410.

[0061] The memory 450 includes volatile memory or nonvolatile memory, or both. Nonvolatile memory can be read only memory (ROM), programmable ROM (PROM), erasable ROM (EPROM), electrically erasable ROM (EEPROM), flash memory, or the like. Volatile memory can include random access memory (RAM), dynamic RAM (DRAM), static RAM (SRAM), fast page mode DRAM (FPM DRAM), extended data output RAM (EDO RAM), extended data output dual data rate RAM (EDO DDR RAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous DRAM (SSDRAM), video random access memory (VRAM), cache memory, or the like.

[0062] In some embodiments, the memory 450 is capable of storing data to support various operations, examples of which include programs, modules, and data structures or a subset or superset thereof, examples of which are illustrated below.

[0063] The operating system 451 includes various procedures, sets of instructions, or code modules for handling various basic system services and for performing hardware dependent tasks, such as firmware and / or hardware initialization, power control, and the like.

[0064] The network communication module 452 is for communicating with other electronic devices via one or more network interfaces 420 (wired or wireless), examples of which include Bluetooth, WiFi, Universal Serial Bus (USB), and the like.

[0065] The presentation module 453 is for enabling presentation of information via one or more output devices 431 (e.g., a display, a speaker, etc.) associated with the user interface 430 (e.g., a user interface for operating a peripheral device and displaying content and information).

[0066] The input processing module 454 is for detecting and interpreting one or more user inputs or interactions from one or more input devices 432.

[0067] In some embodiments, the point of interest processing apparatus provided by the embodiments of the present application can be implemented in software, Figure 2 A point of interest processing apparatus 455 stored in the memory 450 is shown, which can be in the form of software such as programs and plug-ins, including the following software modules: a feature mining module 4551 and a state prediction module 4552, which are logical, and thus can be combined or further split according to the functions implemented. The functions of the various modules will be described below.

[0068] The application will be described in detail below with reference to the exemplary application and implementation of the electronic device provided by the embodiments of the application, and the point of interest processing method provided by the embodiments of the application.

[0069] Referring to Figure 3A , Figure 3A is a flowchart of the point of interest processing method provided by the embodiments of the application, which will be described in detail below with reference to the steps shown in the flowchart. Figure 3A

[0070] In step 101, the parking heat features of road segments respectively passed through by a plurality of historical driving trajectories are determined according to the speeds of the plurality of historical driving trajectories.

[0071] Here, the historical driving trajectory is the driving trajectory of a vehicle in history, and the plurality of historical driving trajectories can correspond to the same vehicle or different vehicles. In order to ensure the accuracy of the point of interest processing, a corresponding constraint condition can be set for the historical driving trajectory, that is, only the historical driving trajectory that meets the constraint condition can be used to predict the state of the point of interest. The constraint condition can include at least one of a constraint area and a constraint time period. Taking a certain point of interest as an example, the constraint area can be a surrounding area centered on the point of interest, and the shape and area of the surrounding area can be set according to the actual application scenario. In this way, the relevance of the historical driving trajectory and the point of interest can be ensured, and the waste of computing resources caused by processing irrelevant historical driving trajectories can be avoided. The constraint time period can be a set time period before the real time, such as the past two weeks before the real time. In this way, the real-time nature of the historical driving trajectory can be ensured, and the real-time nature and effectiveness of the subsequently predicted state can be ensured.

[0072] For each historical driving trajectory obtained, the road segments passed through by the historical driving trajectory and in which parking is performed are identified according to the speed in the historical driving trajectory. For each road segment, the parking heat feature of the road segment is determined according to the number of historical driving trajectories that parked in the road segment, for example, the number of historical driving trajectories that parked in the road segment can be directly used as the parking heat feature of the road segment, or the number of historical driving trajectories that parked in the road segment can be weighted to obtain the parking heat feature of the road segment. The weight used for weighting can be set according to the actual application scenario.

[0073] It is worth noting that the speed in the historical driving trajectory is usually not constant, that is, different positions in the historical driving trajectory can correspond to different speeds, and therefore the parking position in the historical driving trajectory can be identified according to the speed. The road segment is obtained by road segmenting the road in the road network data, and the segmenting distance used for road segmenting can be set according to the actual application scenario, such as 10 meters.​

[0074] In some embodiments, before step 101, the method further comprises: obtaining a plurality of historical trajectory points, wherein each trajectory point corresponds to identification information and a recording time; sorting the plurality of trajectory points corresponding to the same identification information according to the recording time to obtain a historical driving trajectory corresponding to the same identification information; and determining a neighboring distance between each trajectory point and a neighboring trajectory point in the historical driving trajectory, and deleting a trajectory point with a neighboring distance greater than or equal to a neighboring distance threshold.

[0075] Here, a plurality of historical trajectory points can be obtained, each trajectory point corresponding to identification information and a recording time. For example, the trajectory points can be obtained according to the constraint conditions described above. The recording manner of the trajectory points is not limited by the embodiments of the present application. For example, the trajectory points can be recorded by a vehicle machine device in the vehicle, or recorded by a mobile terminal device held by a user in the vehicle. Accordingly, the identification information corresponding to the trajectory points can be a vehicle identification, a device identification of a device used to record the trajectory points, or an account identification of an account in a device used to record the trajectory points.

[0076] After obtaining the plurality of trajectory points, the plurality of trajectory points corresponding to the same identification information are sorted according to the recording time to obtain a historical driving trajectory corresponding to the same identification information. For ease of understanding, a schematic diagram of the trajectory points is provided as shown in Table 1:

[0077] Table 1

[0078] Trajectory point name Identification information Recording time Trajectory point 1 Vehicle 1 January 1, 16:00:00 Trajectory point 2 Vehicle 1 January 1, 16:00:10 Trajectory point 3 Vehicle 2 January 1, 16:00:11 Trajectory point 4 Vehicle 2 January 1, 16:00:16 Trajectory point 5 Vehicle 1 January 1, 16:00:20

[0079] Taking the case where the identification information is a vehicle identification as an example, two historical driving trajectories, i.e., “trajectory point 1->trajectory point 2->trajectory point 5” and “trajectory point 3->trajectory point 4”, can be obtained after sorting.

[0080] For each historical driving trajectory determined, a neighboring distance between each trajectory point and a neighboring trajectory point in the historical driving trajectory is determined, and a trajectory point with a neighboring distance greater than or equal to a neighboring distance threshold is deleted, wherein the neighboring trajectory point can include at least one of a previous trajectory point and a next trajectory point, and the neighboring distance threshold can be set according to an actual application scenario, such as 100 meters. Taking the case where the neighboring trajectory point includes the previous trajectory point and the next trajectory point as an example, in the historical driving trajectory “trajectory point 1->trajectory point 2->trajectory point 5”, when the distance between trajectory point 2 and trajectory point 1 is greater than or equal to the neighboring distance threshold, and the distance between trajectory point 2 and trajectory point 5 is greater than or equal to the neighboring distance threshold, trajectory point 2 is deleted in the historical driving trajectory. In this way, the abnormal trajectory points in the historical driving trajectory are removed, and the effectiveness of the historical driving trajectory is ensured.

[0081] In step 102, according to the distance between the plurality of historical driving trajectories and the interest point, the passing driving trajectory passing through the interest point is screened out from the plurality of historical driving trajectories, and the passing heat feature of the interest point is determined according to the number of the passing driving trajectory.

[0082] Here, for each historical driving trajectory, whether the historical driving trajectory is a passing driving trajectory passing through the interest point is judged according to the distance between the historical driving trajectory and the interest point, wherein the judgment can be made according to a set passing distance threshold (such as 15 meters). Then, the passing heat feature of the interest point is determined according to the number of the screened passing driving trajectory, for example, the number of the passing driving trajectory can be directly taken as the passing heat feature of the interest point, or the number of the passing driving trajectory can also be weighted to obtain the passing heat feature of the interest point, and the weight used for the weighting can be set according to the actual application scenario.

[0083] In step 103, according to the orientation angle similarity between the plurality of passing driving trajectories and the interest point, the passing driving trajectory passing through the interest point is screened out from the plurality of passing driving trajectories, and the passing heat feature of the interest point is determined according to the number of the passing driving trajectory.

[0084] Here, for each passing driving trajectory, whether the passing driving trajectory is a passing driving trajectory passing through the interest point is judged according to the orientation angle similarity between the passing driving trajectory and the interest point, wherein the orientation angle can be a value in the range of [0, 360 degrees], and the form is not limited to this; the screening of the passing driving trajectory can be made according to a set similarity threshold (such as 30 degrees). Then, the passing heat feature of the interest point is determined according to the number of the screened passing driving trajectory, for example, the number of the passing driving trajectory can be directly taken as the passing heat feature of the interest point, or the number of the passing driving trajectory can also be weighted to obtain the passing heat feature of the interest point, and the weight used for the weighting can be set according to the actual application scenario.

[0085] In step 104, the state prediction processing is performed according to the parking heat feature, the passing heat feature and the passing heat feature, and the state for representing whether the interest point is passable is obtained.

[0086] After the feature mining of steps 101-103, the state prediction processing is performed according to the mined parking heat feature, passing heat feature and passing heat feature, and the state of the interest point is obtained, which is used to represent whether the interest point is passable. Since the three kinds of features mined have strong correlation with the state of the interest point, the accuracy of the predicted state can be guaranteed by combining the three kinds of features.

[0087] In this embodiment of the application, a state prediction model can be constructed based on machine learning principles to achieve state prediction processing. The type of state prediction model is not limited; for example, it can be a binary classification model based on a neural network. Figure 4A As shown, the input to the state prediction model can include parking popularity features of multiple road segments, transit popularity features of the point of interest (POI), and passage popularity features of the POI. The output is the state of the POI. Of course, this is just an example and does not constitute a limitation. For instance, the input to the state prediction model could include parking popularity features of road segments associated with the POI, transit popularity features of the POI, and passage popularity features of the POI; or, for another example, parking popularity features of multiple road segments and parking popularity features of road segments associated with the POI could simultaneously be used as input to the state prediction model.

[0088] Taking the case where the input of the state prediction model includes parking popularity features of multiple road segments, transit popularity features of points of interest (POIs), and traversal popularity features of POIs as an example, the state prediction model can be pre-trained. For instance, the state prediction model can perform state prediction processing on the parking popularity features of multiple sample road segments, transit popularity features of sample POIs, and traversal popularity features of POIs to obtain a state characterizing whether a sample POI is traversable (named the comparison state for easy distinction). Then, based on the difference between the comparison state and the sample state of the sample POI, the state prediction model is trained. The trained state prediction model can then achieve accurate state prediction processing. It is worth noting that the naming of the sample road segments and sample POIs is only used to indicate the training stage of the state prediction model; the sample states can be obtained by manual annotation and are the actual states of the sample POIs. Furthermore, the difference between the state to be compared and the sample state is the loss value. This application does not limit the loss function of the state prediction model. For example, it can be the cross-entropy loss function. The training algorithm of the state prediction model includes, but is not limited to, the Gradient Boosting Decisive Tree (GBDT) algorithm.

[0089] In some embodiments, steps 101 to 104 described above can be executed periodically to periodically update the state of the point of interest and ensure the validity of the state.

[0090] In some embodiments, after step 104, the method further includes: when a route planning request for a point of interest is received, performing the following processing: when the state of the point of interest is traversable, performing route planning processing based on the point of interest; when the state of the point of interest is not traversable, determining the parent point of interest, and performing route planning processing based on the child point of interest whose state of the parent point of interest is traversable.

[0091] Here, when receiving a route planning request (or navigation request) for the interest point, differential processing can be performed according to the state of the interest point. For example, when the state of the interest point is passable, route planning processing is performed according to the interest point; when the state of the interest point is impassable, the parent interest point of the interest point is determined, and route planning processing is performed according to the child interest point of the parent interest point with a passable state. When the number of child interest points with a passable state of the parent interest point includes multiple, at least one of the parking heat characteristics of the road segment associated with the child interest point, the passing heat characteristics of the child interest point, and the passing heat characteristics of the child interest point can be used to filter the multiple child interest points, and route planning processing is performed according to the filtered child interest point. The goal of the filtering processing is to filter out the child interest point with the highest heat. In this way, the feasibility of the planned route is ensured on the basis of meeting the route planning demand. It is worth noting that the route planning request for the interest point can be to take the interest point as the starting point, or to take the interest point as the terminal point.

[0092] In some embodiments, after step 104, it further includes: when receiving a route planning request for the parent interest point of the interest point, performing route planning processing according to the child interest point with a passable state of the parent interest point.

[0093] Here, when receiving a route planning request for the parent interest point of the interest point, route planning processing is performed according to the child interest point with a passable state of the parent interest point, thereby ensuring the effectiveness of the planned route. Similarly, when the number of child interest points with a passable state of the parent interest point includes multiple, the child interest point with the highest heat can be filtered out for route planning processing.

[0094] In some embodiments, after step 104, it further includes: when receiving a search request for the interest point or the parent interest point of the interest point, responding to the search request according to the state of the interest point.

[0095] Here, when receiving a search request for the interest point or the parent interest point of the interest point, the search request can be responded to according to the state of the interest point, for example, the state of the interest point is presented. In this way, the user can know whether the interest point is currently passable, avoiding unnecessary consumption of manpower and material resources. At the same time, the search request can also be responded to according to the interest point information of the interest point, which includes at least one of the following: name, location, and present situation (such as whether it is in a business state, an accessible state, or an available state).

[0096] As Figure 3AAs shown, the embodiment of the present application performs feature mining based on historical driving trajectories. Since the historical driving trajectories are easy to obtain and do not need to be collected manually in the field, the implementation cost can be reduced and the processing efficiency can be improved. Meanwhile, the state of the interest point can be predicted by combining the three aspects of the heat features mined, so that the accuracy of the predicted state can be ensured.

[0097] In some embodiments, referring to Figure 3B , Figure 3B is a flowchart of a method for processing an interest point provided by the embodiment of the present application, Figure 3A The step 101 shown can be implemented by the steps 201 to 203, which will be described in combination with the steps.

[0098] In the step 201, the road network data is processed by road segmentation according to a segmentation distance, to obtain a plurality of road segments.

[0099] Here, the road in the road network data can be processed by road segmentation according to a set segmentation distance, to obtain a plurality of road segments. The embodiment of the present application does not limit the way of obtaining the road network data, for example, the road network data can be obtained from a traffic department or an electronic map service provider.

[0100] In the step 202, the active parking position in any one of the historical driving trajectories is determined according to the speed of the plurality of trajectory points in the historical driving trajectory, and a corresponding relationship between the active parking position and the road segment closest to the active parking position is established.

[0101] Here, for each historical driving trajectory, the active parking position in the historical driving trajectory is determined according to the speed of the plurality of trajectory points in the historical driving trajectory, and the distance between the active parking position and the plurality of road segments is calculated, to establish a corresponding relationship between the active parking position and the road segment closest to the active parking position.

[0102] In some embodiments, the active parking position in any one of the historical driving trajectories can be determined according to the speed of the plurality of trajectory points in the historical driving trajectory in the following way: when the speed of at least part of the trajectory points in any one of the historical driving trajectories is zero, and the time length of the records of the at least part of the trajectory points at the same position is greater than or equal to a time length threshold, the same position is determined as a candidate parking position; the passive parking position caused by traffic factors is determined according to the road network data; and when the candidate parking position and the passive parking position fail to match, the candidate parking position is determined as the active parking position in any one of the historical driving trajectories.

[0103] Here, when the speed of at least part of the trajectory points (the number is greater than 1) in the historical driving trajectory is zero, and the recording duration (the holding duration) of the at least part of the trajectory points at the same position is greater than or equal to the duration threshold, the same position is determined as the candidate parking position. For example, the recording time of the trajectory point 3 in a certain historical driving trajectory is 16:00:11 on January 1, the recording time of the trajectory point 4 is 16:00:16 on January 1, and the trajectory point 3 and the trajectory point 4 are located at the same position. Therefore, the recording duration of the same position is 5 seconds.

[0104] For the determined candidate parking position, it can be directly used as the active parking position. In the embodiment of the present application, the candidate parking position can also be further confirmed. For example, the passive parking position caused by the traffic factor is determined according to the road network data, and the candidate parking position and the passive parking position are matched (that is, whether they are the same position is judged). When the candidate parking position and the passive parking position are successfully matched, it is proved that the candidate parking position is the passive parking position, and therefore no processing is performed. When the candidate parking position and the passive parking position are not matched, the candidate parking position is determined as the active parking position in the historical driving trajectory. The traffic factor can include at least one of the traffic rules (such as the traffic light rule) and the traffic jam (unable to move forward and can only be parked passively). In this way, the effectiveness of the determined active parking position can be ensured, and misrecognition caused by the traffic factor can be avoided.

[0105] In some embodiments, before step 202, the following processing is further included: for any one trajectory point in any one historical driving trajectory, the following processing is performed: determining the adjacent distance between the trajectory point and the adjacent trajectory point; determining the adjacent duration between the recording time of the trajectory point and the recording time of the adjacent trajectory point; and determining the speed of the trajectory point according to the adjacent distance and the adjacent duration.

[0106] For ease of illustration, any one trajectory point in any one historical driving trajectory is taken as an example for illustration. First, the adjacent distance between the trajectory point and the adjacent trajectory point is determined, and the adjacent trajectory point can include at least one of the previous trajectory point and the next trajectory point. The adjacent duration between the recording time of the trajectory point and the recording time of the adjacent trajectory point is determined. The adjacent distance is divided by the adjacent duration to obtain the speed of the trajectory point. In this way, the speed of the trajectory point can be accurately calculated based on the position and the recording time of the trajectory point.

[0107] In step 203, the parking heat feature of any one road segment is determined according to the number of active parking positions corresponding to the road segment.

[0108] Here, for each road segment, a parking heat feature of the road segment is determined according to the number of active parking positions corresponding to the road segment. In this way, the parking heat feature of each road segment can be obtained. It should be noted that the active parking positions need to be distinguished from different historical driving trajectories, for example, historical driving trajectory A and historical driving trajectory B include the same active parking position, but when counting, the number of the active parking position still needs to be counted as 2.

[0109] As shown in Figure 3B , in the case where the historical driving trajectory includes a plurality of trajectory points, the active parking position can be accurately determined according to the speed of the trajectory points, and the parking heat feature of each road segment is further determined.

[0110] In some embodiments, referring to Figure 3C , Figure 3C is a flowchart of a point of interest processing method provided by the embodiments of the present application, Figure 3A The step 102 shown can be implemented by steps 301 to 303, which will be described in combination with each step.

[0111] In step 301, the closest distance between any one historical driving trajectory and the point of interest is determined.

[0112] Here, for each historical driving trajectory, the closest distance between the historical driving trajectory and the point of interest is determined. For example, in the case where the historical driving trajectory includes a plurality of trajectory points, the distances between the plurality of trajectory points in the historical driving trajectory and the point of interest are determined respectively, and the closest distance (the smallest distance) is screened out.

[0113] In step 302, when the closest distance is less than the passing distance threshold, any one historical driving trajectory is determined as a passing driving trajectory passing through the point of interest.

[0114] When the closest distance corresponding to the historical driving trajectory is greater than or equal to the set passing distance threshold, it is proved that the historical driving trajectory does not pass through the point of interest; when the closest distance corresponding to the historical driving trajectory is less than the passing distance threshold, it is proved that the historical driving trajectory passes through the point of interest, and therefore the historical driving trajectory is determined as a passing driving trajectory.

[0115] In step 303, the passing heat feature of the point of interest is determined according to the number of passing driving trajectories.

[0116] Here, the passing heat feature of the point of interest is determined according to the number of passing driving trajectories passing through the point of interest.

[0117] As shown in Figure 3CAs shown, the embodiment of the present application determines whether the historical driving track passes through the interest point according to the nearest distance between the historical driving track and the interest point, and can improve the effectiveness of the determined passing heat feature.

[0118] In some embodiments, referring to Figure 3D , Figure 3D is a flowchart of an interest point processing method provided by the embodiment of the present application, Figure 3A The step 103 shown can be implemented by steps 401 to 404, which will be described in combination with the steps.

[0119] In step 401, the position nearest to the interest point in any one passing driving track is determined as a candidate passing position.

[0120] Here, for each passing driving track passing through the interest point, the position nearest to the interest point in the passing driving track is determined as the candidate passing position. For example, in the case that the passing driving track includes multiple track points, the distances between the multiple track points in the passing driving track and the interest point can be determined respectively, and the position of the track point corresponding to the nearest distance (the smallest distance) is taken as the candidate passing position.

[0121] In step 402, the orientation angle of any one passing driving track at the candidate passing position is determined as the candidate passing orientation angle.

[0122] After the candidate passing position in the passing driving track is determined, the orientation angle of the passing driving track at the candidate passing position is determined as the candidate passing orientation angle. The orientation angle can be recorded together when recording the track point.

[0123] In step 403, when the orientation angle similarity between the candidate passing orientation angle and the orientation angle of the interest point is less than the similarity threshold, any one passing driving track is determined as a passing driving track passing through the interest point.

[0124] Here, the orientation angle similarity between the candidate passing orientation angle and the orientation angle of the interest point is determined. When the orientation angle similarity is greater than or equal to the similarity threshold, no processing is performed; when the orientation angle similarity is less than the similarity threshold, it is proved that the passing driving track corresponding to the candidate passing orientation angle has passed through the interest point, and therefore the passing driving track is determined as the passing driving track.

[0125] In step 404, the passing heat feature of the interest point is determined according to the number of passing driving tracks.

[0126] Here, the passing heat feature of the interest point is determined according to the number of passing driving tracks passing through the interest point.

[0127] As Figure 3DAs shown, the embodiment of the present application can calculate the orientation angle similarity, filter the passing driving track from the multiple driving tracks, and further determine the passing heat feature, so as to ensure the effectiveness of the obtained passing heat feature.

[0128] In some embodiments, referring to Figure 3E , Figure 3E is a flowchart of a point of interest processing method provided by the embodiment of the present application, Figure 3A The step 104 shown can be implemented by steps 501 to 503, which will be described in combination with the steps.

[0129] In step 501, the same-parent parking ratio of the target point of interest is determined according to the parking heat feature of the road segment associated with the target point of interest and the parking heat feature of the road segment associated with the parent point of interest of the target point of interest.

[0130] In the embodiment of the present application, for a point of interest to be predicted (for the sake of distinction, named as a target point of interest), the same-parent parking ratio of the target point of interest can be determined according to the parking heat feature of the road segment associated with the target point of interest and the parking heat feature of the road segment associated with the parent point of interest of the target point of interest, for example, the parking heat feature of the road segment associated with the target point of interest can be divided by the parking heat feature of the road segment associated with the parent point of interest, to obtain the same-parent parking ratio. Wherein, the parking heat features of the road segments respectively associated with all the child points of interest of the parent point of interest can be cumulatively processed (such as summation processing), to obtain the parking heat feature of the road segment associated with the parent point of interest. In this way, the same-parent parking ratio can reflect the importance of the target point of interest among all the child points of interest of the parent point of interest in terms of the parking heat.

[0131] It is worth noting that the embodiment of the present application does not limit the determination manner of the association relationship between the point of interest and the road segment, for example, it can be artificially set, and for example, the association relationship between the point of interest and the nearest road segment can be established.

[0132] In step 502, the same-parent passing ratio of the target point of interest is determined according to the passing heat feature of the target point of interest and the passing heat feature of the parent point of interest of the target point of interest.

[0133] For example, the passing heat features of all child interest points of the parent interest point can be cumulatively processed (such as summation processing), to obtain the passing heat feature of the parent interest point. Then, according to the passing heat feature of the target interest point and the passing heat feature of the parent interest point of the target interest point, the same-parent passing ratio of the target interest point is determined, for example, the passing heat feature of the target interest point can be divided by the passing heat feature of the parent interest point of the target interest point to obtain the same-parent passing ratio. The same-parent passing ratio can reflect the importance of the target interest point among all child interest points of the parent interest point in terms of passing heat.

[0134] In step 503, according to the parking heat features of the plurality of road segments, the passing heat features of the plurality of interest points, the passing heat features of the plurality of interest points, the same-parent parking ratio of the target interest point, and the same-parent passing ratio of the target interest point, state prediction processing is performed to obtain a state for representing whether the target interest point is passable.

[0135] In the embodiments of the present application, since the same-parent parking ratio and the same-parent passing ratio are greatly affected by the actual state of the target interest point, the state prediction processing can be performed according to the parking heat features of the plurality of road segments, the passing heat features of the plurality of interest points (which include the target interest point, for example, the plurality of interest points in a constraint region centered on the target interest point), the passing heat features of the plurality of interest points, the same-parent parking ratio of the target interest point, and the same-parent passing ratio of the target interest point, to obtain a state for representing whether the target interest point is passable. As shown in the figure, the state prediction processing can be implemented by a state prediction model, which is a binary classification model for predicting whether the target interest point is passable. Figure 4B

[0136] In some embodiments, before step 503, the method further includes: performing state prediction processing on the parking heat features of the plurality of sample road segments, the passing heat features of the plurality of sample interest points, the passing heat features of the plurality of sample interest points, the same-parent parking ratio of the target sample interest point, and the same-parent passing ratio of the target sample interest point by the state prediction model to obtain a to-be-compared state for representing whether the target sample interest point is passable; and training the state prediction model according to the difference between the to-be-compared state of the target sample interest point and the sample state.

[0137] ​Here, the state prediction model can be pre-trained to ensure the accuracy of the state prediction processing by the state prediction model. For example, the state prediction model is used to perform state prediction processing on the parking heat features of multiple sample road segments, the passing heat features of multiple sample interest points (for example, multiple interest points in a constraint region centered on the target sample interest point), the passing heat features of multiple sample interest points, the same-parent parking ratio of the target sample interest point (one of the multiple sample interest points), and the same-parent passing ratio of the target sample interest point, to obtain a state (named as a state to be compared for convenience of distinction) representing whether the target sample interest point is passable.

[0138] Then, the difference (loss value) between the state to be compared and the sample state (i.e., the actual state) of the target sample interest point is determined according to the loss function of the state prediction model, and the state prediction model is trained according to the difference. After the state prediction model is trained, the parking heat features of multiple road segments, the passing heat features of multiple interest points, the passing heat features of multiple interest points, the same-parent parking ratio of the target interest point, and the same-parent passing ratio of the target interest point can be processed by the trained state prediction model, so as to improve the accuracy of the state prediction processing.

[0139] As shown in Figure 3E , the same-parent parking ratio and the same-parent passing ratio are further extended based on the mined features, so as to further improve the accuracy of the state prediction processing.

[0140] Next, an exemplary application of the embodiments of the present application in an actual application scenario will be described. The embodiments of the present application provide a scheme for judging the state of a door POI and an entrance POI based on historical driving trajectories, which does not require manual intervention and can judge the state only by a certain amount of historical driving trajectories. The embodiments of the present application can be applied to various application scenarios involving interest points, for example, when planning a route, the interest points with an unpassable state (i.e., closed) can be avoided to ensure that the planned route can be driven (or walked); for example, when receiving a search request, the state of the interest point can be responded to to facilitate the user to know the actual state of the interest point; for example, in the navigation process, the user is reminded of the change of the state of the interest point passed or to be reached, to facilitate switching of the interest point or re-planning of the route to meet the user's navigation needs. For convenience of understanding, the door POI is taken as an example in the following description.

[0141] The input of the embodiment of the application is the name and coordinates of the door POI, the name and coordinates of the parent POI (corresponding to the parent interest point in the foregoing) of the door POI, and the historical driving track in a period of time (corresponding to the constraint condition in the foregoing) near the door POI. The output of the embodiment of the application is the state of the door POI, which is used to represent whether the door POI can be passed through, that is, whether the door POI is open or closed.

[0142] Taking the application to the electronic map APP as an example, the application example of the embodiment of the application is as follows:

[0143] 1) Search result page. When receiving a search request of a user for a door POI or a parent POI of the door POI, the state of the door POI is presented in the search result page of the electronic map APP, for example, a corresponding prompt such as "the door is closed" or "the door cannot be passed through" is presented, to help the user understand the latest state of the door POI.

[0144] 2) Route planning. When receiving a route planning request of a user for a parent POI of a door POI (for example, taking the parent POI as a terminal point), route planning processing is performed for a child POI (corresponding to the child interest point in the foregoing) which is a passable sub-point POI according to the state of the parent POI, so that the validity of the planned route can be ensured, and traffic congestion and impaired user experience caused by guiding the user to a non-passable door POI can be avoided.

[0145] Next, the embodiment of the application will be described from the perspective of bottom-layer implementation. The core of the embodiment of the application is to mine several categories of features with the strongest correlation with the state of the door POI from historical driving tracks, and then perform state prediction processing through a binary classification model. In order to facilitate understanding, a flowchart of an interest point processing method is shown in FIG. 1, and each stage will be described in conjunction with FIG. 2. Figure 5 Figure 5

[0146] 1) Track preprocessing.

[0147] The goal of this stage is to output a standardized track library for use in subsequent stages. The input of this stage is a track point recorded in history, which includes the following fields (attributes): user ID (corresponding to the identification information in the foregoing), track point position (or track point coordinates, in the form of (x, y)), and recording time.

[0148] The steps performed in this stage include:

[0149] ① According to the recording time, sort multiple track points corresponding to the same user ID to obtain a historical driving track corresponding to the same user ID.

[0150] ​​② The historical driving track obtained in step 1 may contain a small amount of track points, and the adjacent distance between any two adjacent track points is greater than the adjacent distance threshold (for example, 100 meters). For example, when the adjacent distance between any track point and the previous track point, and the adjacent distance between the any track point and the next track point are both greater than the adjacent distance threshold, the any track point is regarded as an abnormal track point, and the abnormal track point is deleted.

[0151] ③ The track point in the historical driving track is associated with the nearest road. Here, the track point can also be associated with the nearest road segment.

[0152] ④ For any track point in the historical driving track, the speed of the track point is calculated according to the adjacent distance and the adjacent time length (i.e., the time difference) between the track point and the next track point.

[0153] 2) Parking heat mining.

[0154] The goal of this stage is to output a parking heat feature library. Each data in the parking heat feature library contains a track point, a road associated with the track point, and a total number of parking times within a certain range around the track point.

[0155] The input of this stage is the standardized track library (including a plurality of historical driving tracks) output by the track preprocessing stage. The parking heat mining stage mainly includes the following steps:

[0156] ① Identify the active parking point (i.e., the active parking position) for each historical driving track. For example, when the speed of at least some track points in the historical driving track is zero, and the recording time length (continuous time length) of the at least some track points at the same position is greater than or equal to the time length threshold, the same position is determined as a candidate parking position. On this basis, in combination with the road network data, it is judged whether the candidate parking position is a passive parking position. If not, the candidate parking position is regarded as an active parking position. The passive parking position is a position that is passively parked due to traffic factors, including at least one of traffic rules (such as traffic light rules) and traffic congestion.

[0157] ② Road segmentation is performed on each road according to the segmentation distance, and a plurality of road segments are obtained. The segmentation distance is, for example, 10 meters.

[0158] ③ For each road segment, the total number of times (the number of active parking positions corresponding to the road segment) that all historical driving tracks actively park in the road segment is counted as the parking heat feature of the road segment. Finally, the parking heat feature library is output according to the parking heat features of all road segments.

[0159] 3) Passage heat mining.

[0160] The goal of this stage is to output the passage heat feature library. The data in the passage heat feature library means the number of historical driving trajectories passing by the door POI.

[0161] The input of this stage is the normalized trajectory library output by the trajectory preprocessing stage. The passage heat mining stage includes the following key steps:

[0162] ① Identify historical driving trajectories passing by the door POI. The identification method can be: calculate the nearest distance of each historical driving trajectory to the door POI. If the nearest distance is less than the passage distance threshold (such as 15 meters), it is determined that the corresponding historical driving trajectory passes by the door POI, and the historical driving trajectory is regarded as the passage driving trajectory of the door POI.

[0163] ② Count the number of passage driving trajectories of each door POI (passage heat feature), and finally output the passage heat feature library.

[0164] 4) Passage heat mining.

[0165] The goal of this stage is to output the passage heat feature library. The data in the passage heat feature library means the number of historical driving trajectories passing by the door POI.

[0166] The input of this stage is the passage heat mining stage process data, i.e. the passage driving trajectory of the door POI. The passage heat mining stage includes the following steps:

[0167] ① Determine the orientation angle attribute of the door POI, judge whether the passage driving trajectory has similar orientation (the orientation angle of the trajectory point corresponding to the nearest distance in the passage driving trajectory can be used as the basis for judgment), if it exists, it is determined that the passage driving trajectory passes through the door POI, and the passage driving trajectory is regarded as the passage driving trajectory of the door POI.

[0168] ② Count the number of passage driving trajectories of each door POI (passage heat feature), and finally output the passage heat feature library.

[0169] 5) Door state prediction.

[0170] The goal of this stage is to output the final result: the state of the door POI.

[0171] The input of this stage includes the parking heat feature library output by the parking heat mining stage, the passage heat feature library output by the passage heat mining stage, and the passage heat feature library output by the passage heat mining stage. The door state prediction stage includes the following steps:

[0172] ①Calculate the same-parent parking ratio. For example, the same-parent parking ratio of a door POI = the parking heat feature of the door POI / the sum of parking heat features of all door POIs under the parent point POI of the door POI.

[0173] The same-parent parking ratio is used to reflect the importance of the door POI among all door POIs of the same parent point POI in terms of parking heat. For example, the xx company headquarters building includes a northeast door and a southwest door, and the parking heat feature of the northeast door is greater than that of the southwest door, so the same-parent parking ratio of the northeast door is also greater. For another example, the xx building has only one door POI, and the same-parent parking ratio of the door POI is 1.

[0174] ②Calculate the same-parent passing ratio. For example, the same-parent passing ratio of a door POI = the passing heat feature of the door POI / the sum of passing heat features of all door POIs under the parent point POI of the door POI.

[0175] The same-parent passing ratio is used to reflect the importance of the door POI among all door POIs of the same parent point POI in terms of passing heat. For example: For some multi-entry residential areas, the passing heat feature of the main door is usually greater than that of other doors (such as side doors), so the same-parent passing ratio of the main door is also greater, but not 1; for residential areas with only one entrance, the same-parent passing ratio of the entrance is 1.

[0176] ③Predict the state of the door POI based on a binary classification model. The feature input of the model includes: the parking heat feature library, the passing heat feature library, the passing heat feature library, the same-parent parking ratio of the door POI to be predicted, and the same-parent passing ratio of the door POI. For model selection, various binary classification models can be applied, such as a binary classification model based on GBDT.

[0177] Before actual application, the binary classification model can be pre-trained, and the training set can be artificially produced, covering door POIs with actual state of opening (passable) and door POIs with actual state of closing (unpassable), and the number of orders of magnitude can be one hundred thousand, but is not limited thereto.

[0178] Compared with the scheme provided by the related art, the embodiments of the present application have at least the following technical effects: 1) From the cost-benefit point of view: Since the embodiments of the present application do not involve field collection and do not require investment in artificial production, the cost is extremely low, the cost of the production process is almost 0, and the cost-benefit of applying the embodiments of the present application to predict the state of the door POI is about 5 yuan / POI; 2) From the efficiency benefit point of view, since the embodiments of the present application do not involve time-consuming processes such as field collection and artificial production, and completely rely on program automation production, the update cycle of the state can be shortened to 10 seconds / POI or even shorter, and the production efficiency is greatly improved.

[0179] It can be understood that, in the embodiments of the present application, the data related to the track points, the historical driving trajectories, the interest points, etc. need to be obtained with the permission or consent of the user when the embodiments of the present application are applied to specific products or technologies, and the collection, use and processing of the related data need to comply with the relevant laws, regulations and standards of the country and region.

[0180] The following continues to describe an exemplary structure of the interest point processing apparatus 455 implemented as a software module provided by the embodiments of the present application. In some embodiments, as shown in FIG. 4B, the software module stored in the interest point processing apparatus 455 of the memory 450 can include: Figure 2 As shown in FIG. 4B, the software module stored in the interest point processing apparatus 455 of the memory 450 can include: a feature mining module 4551, configured to: determine a parking heat feature of a road segment through by the plurality of historical driving trajectories according to the speed of the plurality of historical driving trajectories; filter out a through driving trajectory that passes through the interest point from the plurality of historical driving trajectories according to the distance between the plurality of historical driving trajectories and the interest point, and determine a through heat feature of the interest point according to the number of the through driving trajectories; filter out a passing driving trajectory that passes through the interest point from the plurality of through driving trajectories according to the angle similarity between the plurality of through driving trajectories and the interest point, and determine a passing heat feature of the interest point according to the number of the passing driving trajectories; a state prediction module 4552, configured to perform state prediction processing according to the parking heat feature, the through heat feature and the passing heat feature to obtain a state for representing whether the interest point is passable.

[0181] In some embodiments, the historical driving trajectory includes a plurality of track points; and the feature mining module 4551 is further configured to: perform road segmentation processing on the road network data according to a segmentation distance to obtain a plurality of road segments; determine a voluntary parking position in any one of the historical driving trajectories according to the speed of the plurality of track points in the any one of the historical driving trajectories, and establish a corresponding relationship between the voluntary parking position and a road segment closest to the voluntary parking position; and determine a parking heat feature of any one of the road segments according to the number of the voluntary parking positions corresponding to the any one of the road segments.

[0182] In some embodiments, the feature mining module 4551 is further configured to: when the speed of at least part of the track points in any one of the historical driving trajectories is zero, and the record duration of the at least part of the track points at the same position is greater than or equal to a duration threshold, determine the same position as a candidate parking position; determine a passive parking position that is passively parked due to a traffic factor according to the road network data; and when the candidate parking position and the passive parking position fail to match, determine the candidate parking position as the voluntary parking position in the any one of the historical driving trajectories.

[0183] In some embodiments, the feature mining module 4551 is further configured to: for any one trajectory point in any one historical driving trajectory, perform the following processing: determine a neighboring distance between the any one trajectory point and a neighboring trajectory point; determine a neighboring time length between a recording time of the any one trajectory point and a recording time of the neighboring trajectory point; and determine a speed of the any one trajectory point according to the neighboring distance and the neighboring time length.

[0184] In some embodiments, the feature mining module 4551 is further configured to: for any one historical driving trajectory, perform the following processing: determine a nearest distance between the any one historical driving trajectory and a point of interest; and determine the any one historical driving trajectory as a passing driving trajectory passing the point of interest when the nearest distance is less than a passing distance threshold.

[0185] In some embodiments, the feature mining module 4551 is further configured to: for any one passing driving trajectory, perform the following processing: determine a candidate threading position in the any one passing driving trajectory as a position closest to the point of interest; determine a candidate threading orientation angle in the any one passing driving trajectory at the candidate threading position as a threading orientation angle of the any one passing driving trajectory; and determine the any one passing driving trajectory as a threading driving trajectory threading the point of interest when an orientation angle similarity between the candidate threading orientation angle and an orientation angle of the point of interest is less than a similarity threshold.

[0186] In some embodiments, the feature mining module 4551 is further configured to: obtain a plurality of trajectory points in history; each trajectory point corresponds to identification information and a recording time; sort a plurality of trajectory points corresponding to the same identification information according to the recording time to obtain a historical driving trajectory corresponding to the same identification information; and determine a neighboring distance between each trajectory point and a neighboring trajectory point in the historical driving trajectory, and delete a trajectory point with a neighboring distance greater than or equal to a neighboring distance threshold.

[0187] In some embodiments, the state prediction module 4552 is further configured to: determine a same-parent parking ratio of a target point of interest according to a parking heat feature of a road segment associated with the target point of interest and a parking heat feature of a road segment associated with a parent point of interest of the target point of interest; determine a same-parent threading ratio of the target point of interest according to a threading heat feature of the target point of interest and a threading heat feature of the parent point of interest of the target point of interest; and perform state prediction processing according to the parking heat features of the plurality of road segments, the passing heat features of the plurality of points of interest, the threading heat features of the plurality of points of interest, the same-parent parking ratio of the target point of interest, and the same-parent threading ratio of the target point of interest to obtain a state representing whether the target point of interest is threadable.

[0188] In some embodiments, the state prediction module 4552 is further configured to: perform state prediction processing on the parking heat feature of the plurality of sample road segments, the passing heat feature of the plurality of sample interest points, the penetrating heat feature of the plurality of sample interest points, the same-parent parking ratio of the target sample interest point, and the same-parent penetrating ratio of the target sample interest point by using the state prediction model, to obtain a to-be-compared state for representing whether the target sample interest point is penetrable; and train the state prediction model according to a difference between the to-be-compared state of the target sample interest point and the sample state. The trained state prediction model is configured to perform state prediction processing on the parking heat feature of the plurality of road segments, the passing heat feature of the plurality of interest points, the penetrating heat feature of the plurality of interest points, the same-parent parking ratio of the target interest point, and the same-parent penetrating ratio of the target interest point.

[0189] In some embodiments, the interest point processing apparatus 455 further includes a request response module configured to: when receiving a route planning request for an interest point, perform the following processing: when the state of the interest point is penetrable, performing route planning processing according to the interest point; and when the state of the interest point is not penetrable, determining a parent interest point of the interest point, and performing route planning processing according to a child interest point of the parent interest point whose state is penetrable.

[0190] In some embodiments, the request response module is further configured to: when receiving a route planning request for a parent interest point of an interest point, performing route planning processing according to a child interest point of the parent interest point whose state is penetrable.

[0191] In some embodiments, the request response module is further configured to: when receiving a search request for an interest point or a parent interest point of the interest point, responding to the search request according to the state of the interest point.

[0192] The embodiment of the present application provides a computer program product or a computer program, the computer program product or the computer program includes computer instructions (i.e., executable instructions), and the computer instructions are stored in a computer readable storage medium. The processor of the electronic device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the electronic device executes the interest point processing method provided in the embodiment of the present application.

[0193] The embodiment of the present application provides a computer readable storage medium, which stores executable instructions. When the executable instructions are executed by a processor, the processor will execute the interest point processing method provided by the embodiment of the present application.

[0194] In some embodiments, the computer readable storage medium can be FRAM, ROM, PROM, EPROM, EEPROM, flash memory, magnetic surface memory, optical disc, or CD-ROM, etc. memory; or can be various devices including one or any combination of the above memories.

[0195] In some embodiments, executable instructions can be in the form of programs, software, software modules, scripts, or code, written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages; and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0196] By way of example, an executable instruction can be, but is not limited to, a file in a file system, can be stored in a part of a file that holds other programs or data, for example, one or more scripts stored in a web page; in a single file dedicated to the program in question, or in multiple coordinated files, for example, files that store one or more modules, sub programs, or code portions.

[0197] By way of example, an executable instruction can be deployed to execute on one electronic device or on multiple electronic devices that are located at one site, or that are distributed across multiple sites and that are interconnected by a communication network.

[0198] The above merely provides an example of the embodiments of the present application, and is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, and improvement made within the spirit and scope of the present application shall be included in the protection scope of the present application.

Claims

1. A point of interest processing method, characterized by, The method comprises: determining, according to the speed of a plurality of historical driving trajectories, the parking heat characteristics of road segments respectively passed through by the plurality of historical driving trajectories; screening, according to the distance between the plurality of historical driving trajectories and a point of interest, a passing driving trajectory passing through the point of interest from the plurality of historical driving trajectories, and determining the passing heat characteristics of the point of interest according to the number of the passing driving trajectories; screening, according to the angle similarity between a plurality of the passing driving trajectories and the point of interest, a passing-through driving trajectory passing through the point of interest from the plurality of the passing driving trajectories, and determining the passing-through heat characteristics of the point of interest according to the number of the passing-through driving trajectory; dividing the parking heat characteristics of a road segment associated with a target point of interest by the parking heat characteristics of a road segment associated with a parent point of interest of the target point of interest to obtain a same-parent parking ratio of the target point of interest; dividing the passing-through heat characteristics of the target point of interest by the passing-through heat characteristics of the parent point of interest of the target point of interest to obtain a same-parent passing-through ratio of the target point of interest; wherein the passing-through heat characteristics of the parent point of interest is obtained by summing the passing-through heat characteristics of all child points of interest of the parent point of interest; performing state prediction processing on the parking heat characteristics of a plurality of road segments, the passing heat characteristics of the point of interest, the passing-through heat characteristics of the point of interest, the same-parent parking ratio of the target point of interest, and the same-parent passing-through ratio of the target point of interest by using a pre-trained state prediction model to obtain a state for representing whether the target point of interest is passable.

2. The method of claim 1, wherein, The historical driving trajectory comprises a plurality of trajectory points; and the determining, according to the speed of a plurality of historical driving trajectories, the parking heat characteristics of road segments respectively passed through by the plurality of historical driving trajectories comprises: performing road segmentation processing on road network data according to a segmentation distance to obtain the plurality of road segments; determining a voluntary parking position in any one historical driving trajectory according to the speed of a plurality of trajectory points in the any one historical driving trajectory, and establishing a corresponding relationship between the voluntary parking position and a road segment closest to the voluntary parking position; determining the parking heat characteristics of any one road segment according to the number of voluntary parking positions corresponding to the any one road segment.

3. The method of claim 2, wherein, The determining, according to the speed of a plurality of trajectory points in any one historical driving trajectory, the voluntary parking position in the any one historical driving trajectory comprises: when the speed of at least part of the trajectory points in the any one historical driving trajectory is zero, and the recording time length of the at least part of the trajectory points at the same position is greater than or equal to a time length threshold, determining the same position as a candidate parking position; determining a passive parking position in which a vehicle is passively parked due to a traffic factor according to the road network data; when the candidate parking position and the passive parking position fail to match, determining the candidate parking position as the voluntary parking position in the any one historical driving trajectory.

4. The method of claim 2, wherein, Before the determining, according to the speed of a plurality of trajectory points in any one historical driving trajectory, the voluntary parking position in the any one historical driving trajectory, the method further comprises: For any one of the trajectory points in any one of the historical driving trajectories, the following processing is performed: determine the adjacent distance between the arbitrary trajectory point and the adjacent trajectory point; determine the adjacent time length between the recording time of the arbitrary trajectory point and the recording time of the adjacent trajectory point; determine the speed of the arbitrary trajectory point according to the adjacent distance and the adjacent time length.

5. The method of claim 1, wherein, The method further comprises: According to the distance between the plurality of historical driving trajectories and the interest point, the passing driving trajectory passing through the interest point is screened out from the plurality of historical driving trajectories, comprising: For any one historical driving trajectory, the following processing is performed: determine the nearest distance between the arbitrary historical driving trajectory and the interest point; 6. The method of claim 1, wherein, when the nearest distance is less than the passing distance threshold, the arbitrary historical driving trajectory is determined as the passing driving trajectory passing through the interest point. The method further comprises: According to the direction angle similarity between a plurality of the passing driving trajectories and the interest point, the passing driving trajectory passing through the interest point is screened out from a plurality of the passing driving trajectories, comprising: For any one passing driving trajectory, the following processing is performed: determine the position closest to the interest point in the arbitrary passing driving trajectory as the candidate passing position; 7. The method of claim 1, wherein, determine the direction angle of the arbitrary passing driving trajectory at the candidate passing position as the candidate passing direction angle; when the direction angle similarity between the candidate passing direction angle and the direction angle of the interest point is less than the similarity threshold, the arbitrary passing driving trajectory is determined as the passing driving trajectory passing through the interest point. Before the method further comprises: obtain a plurality of trajectory points in history; wherein each of the trajectory points corresponds to identification information and recording time; 8. The method of claim 1, wherein, sort a plurality of trajectory points corresponding to the same identification information according to the recording time to obtain a historical driving trajectory corresponding to the same identification information; determine the adjacent distance between each trajectory point and the adjacent trajectory point in the historical driving trajectory, and delete the trajectory point with an adjacent distance greater than or equal to the adjacent distance threshold. The method further comprises: perform state prediction processing on the parking heat features of a plurality of sample road segments, the passing heat features of a sample interest point, the passing heat features of the sample interest point, the same parent parking ratio of a target sample interest point, and the same parent passing ratio of the target sample interest point through a state prediction model to obtain a to-be-compared state for representing whether the target sample interest point is passable; 9. The method according to any one of claims 1 to 8, characterized in that, train the state prediction model according to the difference between the to-be-compared state of the target sample interest point and the sample state; wherein the trained state prediction model is used for state prediction processing on the parking heat features of the plurality of road segments, the passing heat features of the interest point, the passing heat features of the interest point, the same parent parking ratio of the target interest point, and the same parent passing ratio of the target interest point. The method further comprises: When a route planning request for the interest point is received, the following processing is performed: When the state of the interest point is passable, route planning processing is performed according to the interest point; When the state of the interest point is impassable, a parent interest point of the interest point is determined, and route planning processing is performed according to a child interest point of the parent interest point whose state is passable.

10. The method according to any one of claims 1 to 8, characterized in that, The method further includes: When a route planning request for the parent interest point of the interest point is received, route planning processing is performed according to a child interest point of the parent interest point whose state is passable.

11. The method according to any one of claims 1 to 8, characterized in that, The method further includes: When a search request for the interest point or the parent interest point of the interest point is received, the search request is responded to according to the state of the interest point.

12. A point of interest processing apparatus characterized by comprising: The device includes: A feature mining module configured to: Determine a parking heat feature of a road segment through which a plurality of historical driving trajectories respectively pass according to speeds of the plurality of historical driving trajectories; Filter out a passing driving trajectory that passes through an interest point from the plurality of historical driving trajectories according to distances between the plurality of historical driving trajectories and the interest point, and determine a passing heat feature of the interest point according to a number of the passing driving trajectories; Filter out a passing-through driving trajectory that passes through the interest point from a plurality of the passing driving trajectories according to a similarity of orientation angles between the plurality of the passing driving trajectories and the interest point, and determine a passing-through heat feature of the interest point according to a number of the passing-through driving trajectories; A state prediction module configured to divide a parking heat feature of a road segment associated with a target interest point by a parking heat feature of a road segment associated with a parent interest point of the target interest point to obtain a same-parent parking ratio of the target interest point; Divide a passing-through heat feature of the target interest point by a passing-through heat feature of the parent interest point of the target interest point to obtain a same-parent passing-through ratio of the target interest point, wherein the passing-through heat feature of the parent interest point is obtained by summing passing-through heat features of all child interest points of the parent interest point; Perform state prediction processing on the parking heat features of the plurality of road segments, the passing heat feature of the interest point, the passing-through heat feature of the interest point, the same-parent parking ratio of the target interest point, and the same-parent passing-through ratio of the target interest point by using a pre-trained state prediction model to obtain a state for representing whether the target interest point is passable.

13. The apparatus of claim 12, wherein, The feature mining module is further configured to: Perform road segmenting processing on road network data according to a segmenting distance to obtain the plurality of road segments; Determine a voluntary parking position in an arbitrary historical driving trajectory according to speeds of a plurality of trajectory points in the arbitrary historical driving trajectory, and establish a corresponding relationship between the voluntary parking position and a road segment closest to the voluntary parking position; Determine a parking heat feature of an arbitrary road segment according to a number of voluntary parking positions that have the corresponding relationship with the arbitrary road segment.

14. The apparatus of claim 13, wherein, The feature mining module is further configured to: determining a candidate parking position when speed of at least part of the trajectory points in the arbitrary historical driving trajectory is zero and a recording time length of the at least part of the trajectory points at a same position is greater than or equal to a time length threshold value; determining a passive parking position caused by a traffic factor according to the road network data; determining the candidate parking position as an active parking position in the arbitrary historical driving trajectory when the candidate parking position fails to match the passive parking position.

15. An electronic device, comprising: The electronic device comprises: a memory for storing executable instructions; a processor for executing the executable instructions stored in the memory to implement the point of interest processing method in any one of claims 1 to 11.

16. A computer-readable storage medium storing executable instructions, wherein the instructions, when executed by a processor, cause the processor to perform operations comprising: The executable instructions are executed by the processor to implement the point of interest processing method in any one of claims 1 to 11.

17. A computer program product comprising computer programs or executable instructions, characterized in that, The computer program or executable instructions are executed by the processor to implement the point of interest processing method in any one of claims 1 to 11.

Images