A vehicle positioning method and system for road network holographic perception

Abstract

The application provides a vehicle positioning method and system for road network holographic perception, and relates to the technical field of data processing. The method comprises correlating and integrating vehicle perception data collected by a roadside terminal, inspection data reported by an inspection personnel, and parking order data obtained, correcting the fusion positioning information by using preset time constraints and space constraints, and mapping the fusion positioning information to a preset digital road network structure model to determine the actual parking berth position of the vehicle in the digital road network. The fusion positioning information is corrected by using the preset time constraints and space constraints and is mapped to the digital road network structure model, so that the actual parking berth position of the vehicle in the digital road network can be accurately determined, the influence caused by the time difference and the space positioning difference between the data from different sources is effectively reduced, the accuracy of the determination of the parking position of the vehicle is improved, and the problem that the parking position of the vehicle is difficult to accurately determine is overcome.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a vehicle positioning method and system for road network holographic perception. Background Technology

[0002] With the continuous improvement of urban digital governance, digital road network systems based on refined management of road resources are gradually becoming an important infrastructure for urban traffic management and smart parking management. By deploying various sensing devices along roads and combining them with digital maps, dynamic perception and management of road resources, vehicle status, and parking space usage can be achieved, thus providing data support for urban traffic scheduling, parking supervision, and road resource utilization. In this process, accurately obtaining the actual parking location of vehicles in road parking spaces within the digital road network environment has become a crucial technological foundation for supporting road network perception and parking management.

[0003] Vehicle location is typically identified and inferred using a single type of sensor data. For example, roadside video equipment or geomagnetic detection equipment can be used to detect vehicle entry and exit, combined with order information or vehicle identification information from a parking management platform to determine whether a vehicle is parked. Alternatively, manual inspections can be conducted to record vehicle parking information, which is then compared in the background to manage vehicle parking behavior. These methods can, to a certain extent, obtain information on vehicle parking within road areas, thus providing basic data support for parking management.

[0004] Although vehicle parking status can be identified through single sensor data or simple data comparison, in actual road environments, due to differences in collection time, collection location, and data accuracy between different data sources, it is easy to encounter the problem of difficulty in uniformly determining the parking location when making comprehensive judgments on multi-source information of the same vehicle. Summary of the Invention

[0005] To overcome the problem of accurately determining the parking location of vehicles, this application provides a vehicle positioning method and system based on road network holographic perception.

[0006] On one hand, the present invention provides a vehicle positioning method based on road network holographic perception, comprising: collecting vehicle license plate information through a roadside terminal and recording the appearance and disappearance times of the corresponding vehicles to construct vehicle perception data; receiving vehicle license plate information, inspection verification time, and inspection location reported by inspection personnel to construct inspection data; acquiring vehicle license plate information, order generation time, payment time or cancellation time, and road segment identifier and parking space identifier associated with the parking order from parking orders to construct parking order data; associating and integrating the vehicle perception data, the inspection data, and the parking order data to generate fused positioning information; correcting the fused positioning information through preset time constraints and spatial constraints to obtain corrected fused positioning information; and mapping the corrected fused positioning information to a preset digital road network structure model to determine the actual parking space location of the vehicle in the digital road network.

[0007] Optionally, the step of collecting vehicle license plate information through a roadside terminal and recording the appearance and disappearance times of the corresponding vehicles to construct vehicle perception data includes: continuously acquiring road images within the monitoring area using the image acquisition module built into the roadside terminal to obtain a continuous frame image sequence; performing vehicle target detection on each frame of the continuous frame image sequence to extract candidate image blocks of the license plate area, and performing character recognition on the candidate image blocks to obtain the license plate information of the corresponding vehicle; recording the system timestamp corresponding to the first recognition of the license plate information as the appearance time of the vehicle; continuously tracking and detecting the continuous frame image sequence, and when no vehicle corresponding to the license plate information is detected within a preset number of consecutive frames, recording the system timestamp corresponding to the last successful recognition of the license plate information as the disappearance time of the vehicle; associating and encapsulating the license plate information, the appearance time, the disappearance time, and the device identifier of the roadside terminal to generate vehicle perception data for the corresponding vehicle.

[0008] Optionally, the step of receiving vehicle license plate information, inspection verification time, and inspection location reported by inspection personnel to construct inspection data includes: having inspection personnel use an inspection terminal to conduct on-site inspections of parked vehicles along a preset inspection route; using the image acquisition module built into the inspection terminal to photograph the license plate area of ​​the parked vehicles and performing character recognition on the photographed images to obtain the license plate information of the corresponding vehicles; obtaining the current system timestamp as the inspection verification time and calling the positioning module built into the inspection terminal to obtain the current geographical coordinates as the inspection location; and associating and encapsulating the license plate information, the inspection verification time, the inspection location, and the inspection personnel's employee ID to form the inspection data for the corresponding vehicles.

[0009] Optionally, the step of obtaining the vehicle's license plate information, order generation time, payment time or cancellation time, and the road segment identifier and parking space identifier associated with the parking order to construct parking order data includes: triggering an order generation event by the corresponding parking space sensing device when the vehicle enters the parking space, recording the order generation time, the parking space identifier bound to the parking space, and the road segment identifier of the road segment to which the parking space belongs, thus forming a parking order; recording the payment time of the corresponding order when the vehicle completes payment and leaves; recording the cancellation time of the corresponding order when the vehicle has not completed payment and the order exceeds a preset time limit; and associating and encapsulating the order generation time, payment time or cancellation time, road segment identifier, parking space identifier, and the vehicle's corresponding license plate information to generate parking order data for the corresponding vehicle.

[0010] Optionally, the step of associating and integrating the vehicle perception data, the inspection data, and the parking order data to generate fused positioning information includes: using a pre-established parking space number as a spatial anchor point, and based on a preset correspondence between the monitoring area of ​​the roadside terminal and the parking space number, mapping and matching the device identifier in the vehicle perception data, the inspection location in the inspection data, and the parking space identifier in the parking order data with the corresponding parking space number to obtain matched multi-source data; using the order generation time of the parking order data as a benchmark, aligning the matched multi-source data according to the time dimension, incorporating the appearance time and disappearance time in the vehicle perception data and the inspection verification time in the inspection data into a unified time axis to obtain time-aligned multi-source data; and integrating the matched multi-source data and the time-aligned multi-source data to generate fused positioning information.

[0011] Optionally, the step of correcting the fused positioning information through preset time and space constraints to obtain corrected fused positioning information includes: taking the time range in which the inspection verification time falls within the time interval formed by the appearance time and disappearance time of the corresponding vehicle perception data, and the difference between the inspection verification time and the order generation time of the parking order data does not exceed a preset time threshold as a time constraint; taking the spatial range in which the distance between the inspection location in the inspection data and the parking space coordinates corresponding to the parking space identifier in the parking order data does not exceed a preset spatial threshold, and the spatial range in which the monitoring area coverage of the roadside terminal and the parking space location corresponding to the parking space identifier have a spatial overlap relationship as a spatial constraint; using the time constraint and the spatial constraint to verify each record in the fused positioning information in sequence, retaining the records that simultaneously meet the time constraint and the spatial constraint, and updating the retained records as corrected fused positioning information; and marking records that do not meet any constraint rule as abnormal and triggering a manual review process.

[0012] Optionally, the step of mapping the corrected fused positioning information to a preset digital road network structure model to determine the actual parking space location of the vehicle in the digital road network includes: using the parking space number in the corrected fused positioning information as a search key, searching in the preset digital road network structure model based on the search key to obtain the road segment identifier and parking space information corresponding to the parking space number; binding the license plate information in the corrected fused positioning information with the retrieved road segment identifier and parking space information to generate the actual parking space location, wherein the actual parking space location includes the license plate information, the road segment identifier, the specific parking space number, and the parking space coordinates; and inputting the actual parking space location into the corresponding node of the digital road network structure model to realize the visual representation of the actual parking space location of the vehicle in the digital road network.

[0013] On the other hand, this application also provides a vehicle positioning system for road network holographic perception, including: a roadside acquisition module, used to acquire vehicle license plate information through a roadside terminal and record the appearance and disappearance times of the corresponding vehicles to construct vehicle perception data; a personnel acquisition module, used to receive vehicle license plate information, inspection verification time, and inspection location reported by inspection personnel to construct inspection data; an order acquisition module, used to acquire vehicle license plate information, order generation time, payment time or cancellation time, and road segment identifier and parking space identifier associated with the parking order in a parking order to construct parking order data; a data integration module, used to integrate the vehicle perception data, the inspection data, and the parking order data to generate fused positioning information; a constraint correction module, used to correct the fused positioning information through preset time constraints and spatial constraints to obtain corrected fused positioning information; and a data mapping module, used to map the corrected fused positioning information to a preset digital road network structure model to determine the actual parking space location of the vehicle in the digital road network.

[0014] The vehicle positioning method and system for road network holographic perception provided in this application have the following technical effects: Vehicle perception data is constructed by collecting vehicle license plate information and recording the appearance and disappearance times of the corresponding vehicles through roadside terminals; inspection data is constructed by collecting vehicle license plate information, inspection verification time, and inspection location reported by inspection personnel; parking order data is constructed by combining the order generation time, payment time, or cancellation time of parking orders with the associated road segment and parking space identifiers; based on this, vehicle perception data, inspection data, and parking order data are correlated and integrated to generate fused positioning information; and the fused positioning information is corrected through preset time and space constraints. The corrected fused positioning information is then mapped onto a digital road network structure model, thereby accurately determining the actual parking space location of vehicles in the digital road network. This effectively reduces the impact of differences in collection time and spatial positioning between data from different sources, improves the accuracy of vehicle parking location determination, and overcomes the problem of difficulty in accurately determining vehicle parking locations. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating a vehicle positioning method based on road network holographic perception provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the vehicle license plate recognition process provided in an embodiment of the present invention; Figure 3 This is a schematic block diagram illustrating the time constraints provided in an embodiment of the present invention; Figure 4 This is a schematic block diagram illustrating spatial constraints provided in an embodiment of the present invention; Figure 5 This is a schematic block diagram illustrating the structure of the time constraint and space constraint judgment process provided in the embodiments of the present invention; Figure 6 This is a schematic block diagram of the digital road network structure model provided in the embodiments of the present invention; Figure 7 This is a schematic block diagram of a vehicle positioning system for road network holographic perception provided in an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0018] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or sub-modules is not necessarily limited to those steps or sub-modules explicitly listed, but may include other steps or sub-modules not explicitly listed or inherent to such processes, methods, products, or devices.

[0019] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0020] like Figure 1 As shown, this application provides a vehicle localization method based on road network holographic perception, which can determine the actual parking location of a vehicle in a road parking space in an urban road parking management scenario. The method includes: Step S1: Collect vehicle license plate information through roadside terminals and record the appearance and disappearance times of the corresponding vehicles to construct vehicle perception data.

[0021] Roadside terminals are fixedly installed on light poles or dedicated brackets along the side of the road and are bound to a pre-registered road monitoring area via device identification. The roadside terminal continuously identifies vehicles passing through or parked within the monitoring area, extracts vehicle license plate information, and records the time information when a vehicle enters and leaves the monitoring area. The vehicle's appearance time is determined by the system timestamp when the vehicle's license plate information is first identified, and the vehicle's disappearance time is determined by the system timestamp corresponding to the point when the vehicle's license plate information is no longer identified during continuous monitoring. When generating a vehicle record, the roadside terminal associates and encapsulates the license plate information, appearance time, disappearance time, and the roadside terminal's device identification according to a unified data structure to form a complete vehicle perception data record.

[0022] For example, a roadside terminal with the device identifier XXX-1 is installed on road section A. When a vehicle with the license plate information "Yue BXXXXX" enters the monitoring area of the roadside terminal, the image recognition module first recognizes the license plate information at the timestamp 20XX-06-18 09:12:08, and this timestamp is recorded as the appearance time. After the vehicle leaves the monitoring area, when the license plate information "Yue BXXXXX" is no longer recognized in consecutive image frames, and the last recognition result is recorded at the timestamp 20XX-06-18 09:37:46, this timestamp is recorded as the disappearance time. Subsequently, the license plate information "Yue BXXXXX", the appearance time 20XX-06-18 09:12:08, the disappearance time 20XX-06-18 09:37:46 and the device identifier XXX-1 are combined to generate a vehicle sensing data record.

[0023] As shown Figure 2 , in another embodiment, step S1 can be implemented in the following manner: The built-in image acquisition module of the roadside terminal continuously collects road images within its monitoring area to obtain a sequence of consecutive frame images.

[0024] The image acquisition module is an industrial camera device, installed inside the roadside terminal, with the lens facing the road parking area. The preset frame rate is the number of image frames collected per unit time, which is set based on the traffic density of road vehicles, recognition accuracy requirements and device processing capabilities. The continuously collected images are stored as a sequence of consecutive frame images in chronological order, and each frame of image is attached with a corresponding acquisition timestamp for subsequent recognition and time recording.

[0025] For example, in an area with dense road parking spaces, the preset frame rate is set to 10 frames per second, that is, the image acquisition module collects one frame of image every 0.1 seconds, generates 600 frames of image data within 1 minute, and forms a sequence of consecutive frame images in chronological order, with each frame of image attached with a timestamp accurate to the millisecond level.

[0026] Vehicle target detection is performed on each frame of image in the sequence of consecutive frame images to extract the candidate image block of the license plate area, and character recognition is performed on the candidate image block to obtain the license plate information of the corresponding vehicle.

[0027] The vehicle target detection uses a target detection algorithm, for example, the YOLOv5 vehicle detection algorithm based on a convolutional neural network or the Faster R-CNN vehicle detection algorithm, to detect the vehicle target in each frame of image and obtain a vehicle target region. Subsequently, license plate positioning processing is performed within the vehicle target region to extract candidate image blocks of the license plate region. Character recognition processing is performed on the candidate image blocks, and the character recognition processing adopts an OCR character recognition algorithm, for example, the CRNN character recognition algorithm based on a convolutional neural network or the LPRNet license plate recognition algorithm based on deep learning, so as to obtain the complete character sequence of the vehicle license plate as the license plate information.

[0028] For example, two vehicle target regions are detected in one frame of image, the two vehicle regions are respectively extracted by the YOLOv5 vehicle detection algorithm, the license plate region is located in each vehicle region, and then the license plate information "Shanghai AXXXXX" and "Zhejiang BXXXXX" are obtained through recognition by the CRNN character recognition algorithm.

[0029] The system timestamp corresponding to the first time the license plate information is recognized is recorded as the appearance time of the vehicle.

[0030] The roadside terminal maintains a vehicle recognition cache table. When a certain piece of license plate information is recognized for the first time in a continuous frame image sequence, the license plate information is written into the vehicle recognition cache table, and the system timestamp corresponding to the current image frame is recorded as the appearance time. The system timestamp comes from the internal clock module of the roadside terminal, and the clock module is time-calibrated through a network time synchronization protocol to ensure the consistency of time records.

[0031] For example, the license plate information "Guangdong CXXXXX" is recognized for the first time in the 152nd frame of the continuous frame image sequence, and the timestamp of this frame of image is 20XX-07-01 08:15:2X.312, then the appearance time of the license plate information "Guangdong CXXXXX" is recorded in the vehicle recognition cache table as 20XX-07-01 08:15:2X.312.

[0032] Tracking detection is continuously performed on the continuous frame image sequence. When the vehicle corresponding to the license plate information is not detected again within a preset number of consecutive frames, the system timestamp corresponding to the last successful recognition of the license plate information is recorded as the disappearance time of the vehicle.

[0033] The roadside terminal continuously tracks the already recognized license plate information, re-executes vehicle detection and license plate recognition processing in each frame of image, and determines whether the license plate information exists in the recognition result. When the license plate information is not recognized again within the preset number of consecutive frames, it is determined that the vehicle has left the monitoring area. The preset number of frames is the number of consecutive unrecognized frames for judging that the vehicle has left, and this value is set according to the driving speed of road vehicles and the width of the camera's field of view.

[0034] For example, under the condition that the preset frame rate is 10 frames per second, the preset number of frames is set to 30 frames. When the license plate information "Yue CXXXXX" is not recognized again in 30 consecutive frames of images, the timestamp when the license plate information is last recognized is recorded as the disappearance time of the vehicle, for example, recorded as 20XX-07-01 08:18:05.214.

[0035] The license plate information, appearance time, disappearance time and the device identifier of the roadside terminal are associated and encapsulated to generate vehicle perception data corresponding to the vehicle.

[0036] The roadside terminal establishes a unified data structure for the identification record of each vehicle, and writes the license plate information, appearance time, disappearance time and device identifier into the data structure. The device identifier is the unique number of the roadside terminal registered in the device management platform, which is used to identify the specific device from which the vehicle data originates. The data structure is encapsulated in a key-value pair format, and uploaded to the data processing platform through a wireless network for storage.

[0037] For example, a vehicle perception data record may adopt the following structure: {License plate information: "Yue CXXXXX", Appearance time: "20XX-07-01 08:15:2X.312", Disappearance time: "20XX-07-01 08:18:05.214", Device identifier: "XXX-1"}.

[0038] Step S2: Receive the vehicle license plate information, inspection verification time and inspection position reported by the inspection personnel, so as to construct inspection data.

[0039] Inspection personnel perform on-site inspection tasks in the road parking area to manually check parked vehicles. The inspection personnel register the parked vehicles through the portable inspection terminal, and upload the vehicle license plate information, inspection verification time and inspection position to the data processing platform through the inspection terminal. The inspection verification time is automatically generated by the time recording module of the inspection terminal, and the inspection position is acquired by the positioning module of the inspection terminal. When generating an inspection record, the inspection terminal associates and encapsulates the license plate information, inspection verification time, inspection position and the identity identifier of the inspection personnel to form inspection data.

[0040] For example, an inspection personnel with the inspection personnel number 001 conducts on-site verification of road parking spaces at 20XX-07-01 09:02:17, recognizes the vehicle license plate information "Hu AXXXXX" in the inspection terminal, the inspection terminal records the inspection verification time as 20XX-07-01 09:02:17, and acquires the inspection position as longitude 12X.481253 and latitude 3X.223541 through the positioning module, and then generates an inspection data record.

[0041] In another embodiment, step S2 can be implemented in the following manner: Patrol personnel use patrol terminals to conduct on-site inspections of parked vehicles along a pre-set patrol route. The patrol terminals utilize their built-in image acquisition modules to capture images of the license plate areas of parked vehicles and perform character recognition on the captured images to obtain the corresponding vehicle's license plate information.

[0042] The inspection terminal is a mobile device with image acquisition and recognition capabilities. The image acquisition module is used to acquire images of vehicle license plate areas. After acquiring the images, the inspection terminal performs license plate recognition processing. The license plate recognition processing uses license plate recognition algorithms, such as the LPRNet license plate recognition algorithm based on deep learning or the license plate character recognition algorithm based on convolutional neural networks, to recognize characters in the license plate image and obtain complete license plate information.

[0043] For example, inspection personnel take pictures of the license plates of parked vehicles using the inspection terminal. After the inspection terminal recognizes the license plate image, it obtains the license plate information "Su EXXXXX".

[0044] While completing the license plate information recognition, the inspection terminal automatically obtains the current system timestamp as the inspection verification time, and calls the positioning module built into the inspection terminal to obtain the current geographical coordinates as the inspection location.

[0045] The inspection terminal has a built-in time recording module and a positioning module. The time recording module generates the inspection verification time, and the positioning module uses satellite positioning technology, specifically GPS or BeiDou satellite positioning technology, to obtain the inspection location. The positioning module outputs longitude and latitude coordinates to describe the inspection location.

[0046] For example, while recognizing the license plate information "Su EXXXXX", the inspection terminal records the inspection verification time as 20XX-07-01 09:18:45, and obtains the inspection location as longitude 11X.793615 and latitude 3X.060255 through Beidou satellite positioning technology.

[0047] The inspection terminal associates and encapsulates license plate information, inspection verification time, inspection location, and the employee ID of the inspection personnel to form the inspection data for the corresponding vehicle.

[0048] The inspection terminal collects license plate information, inspection verification time, inspection location, and the inspector's employee ID, and writes these into a unified data structure to form an inspection data record. The inspector's employee ID is a unique number for the inspector in the inspection management platform, used to identify the source of the inspection record.

[0049] For example, an inspection data record uses the following structure: {License plate information: "Su EXXXXX", inspection verification time: "20XX-07-01 09:18:45", inspection location: "11X.793615, 3X.060255", inspector employee ID identifier: "001"}.

[0050] Step S3: Acquire the license plate information of the vehicle in the parking order, the order generation time, the payment time or cancellation time, as well as the road section identifier and parking space identifier associated with the parking order, so as to construct parking order data.

[0051] Parking order records related to vehicle parking behaviors are acquired through the road parking charging platform, and the order generation time, payment time or cancellation time, as well as the road section identifier and parking space identifier associated with the parking order are extracted from the parking order records, and the vehicle license plate information bound to the parking order is extracted at the same time. The order generation time is used to represent the time node when the vehicle enters the parking space and triggers the billing event, the payment time is used to represent the time node when the vehicle completes the payment of the parking fee, and the cancellation time is used to represent the time node when the order is automatically closed without completing the payment. By uniformly sorting out the time information, spatial identifier information and vehicle license plate information in the parking order records, they are saved as a complete parking order data record.

[0052] For example, on the parking space P-XXX-05 of road section R-XXX, when a vehicle with the license plate information "Yue BXXXXX" enters the parking space and triggers a billing record, the platform generates the order number 20XX06180015, records the order generation time as 20XX-06-18 09:12:20, and simultaneously records the road section identifier associated with the order as R-XXX and the parking space identifier as P-XXX-05. When the vehicle completes the payment and leaves at 20XX-06-18 09:38:10, the platform records the payment time as 20XX-06-18 09:38:10. Subsequently, the license plate information "Yue BXXXXX", the order generation time 20XX-06-18 09:12:20, the payment time 20XX-06-18 09:38:10, the road section identifier R-XXX and the parking space identifier P-XXX-05 are combined to generate a parking order data record.

[0053] Wherein, the inspection location can be automatically acquired by the positioning module of the inspection terminal, and the positioning module includes a satellite positioning module or a positioning module based on a mobile communication network, which is used to record the current geographic location of the inspection terminal in real time when the inspector performs the vehicle verification operation, and report the geographic location as the inspection location. Through the above method, the inspection location can be consistent with the spatial location where the actual inspection behavior occurs, so as to improve the spatial reference accuracy of the inspection data in the subsequent fusion positioning process.

[0054] For example, when inspection personnel check parked vehicles on a certain road section, the inspection terminal automatically obtains the current location coordinates while recording the vehicle license plate information, and uploads the location coordinates and inspection verification time to the system platform, thereby forming an inspection data record containing license plate information, inspection verification time and inspection location.

[0055] In another embodiment, step S3 can be implemented in the following manner: When a vehicle enters a parking space, the parking management system triggers an order generation event through the corresponding parking space sensing device. It records the order generation time, the parking space identifier bound to the parking space, and the road segment identifier of the road segment to which the parking space belongs, thus forming a parking order.

[0056] Parking space sensing devices are installed in roadside parking areas to detect whether vehicles have entered parking spaces. These devices can be geomagnetic detectors or video detectors. When a vehicle enters a parking space, the sensing device detects that the parking space status has changed from "vacant" to "occupied" and sends an occupancy signal to the parking management platform. Upon receiving the occupancy signal, the parking management platform generates a corresponding parking order, records the order generation time, and simultaneously queries a pre-established parking space information table based on the device number of the parking space sensing device. From the table, it retrieves the parking space identifier corresponding to that device number and the road segment identifier of the parking space.

[0057] For example, a geomagnetic detection device with device number XXX-PXX03 is installed at parking space number P-XXX-03. When a vehicle enters the parking space, the device detects a change in the magnetic field and sends an occupancy signal to the platform. The platform generates a parking order 20XX0701100815 at timestamp 20XX-07-01 10:08:15. At the same time, it queries the parking space information table to find that the parking space identifier corresponding to the device is P-XXX-03 and the road segment identifier is R-XXXX, and writes this information into the parking order record.

[0058] When a vehicle completes payment and leaves, the parking management system records the payment time for the corresponding order.

[0059] After the vehicle completes the parking fee payment, the payment platform sends a payment success message to the parking management platform. Upon receiving the payment success notification, the parking management platform records the payment success time as the payment time and updates the parking order status to "Payment Completed". The payment time is derived from the transaction completion time returned by the payment platform and is stored in conjunction with the parking order number.

[0060] For example, a vehicle with the license plate information "Shanghai AXXXXX" completes payment via mobile payment after finishing parking, the payment platform returns a payment success notification at 20XX-07-01 10:36:22, and the parking management platform records this time as the payment time 20XX-07-01 10:36:22 of the order 20XX0701100815.

[0061] When the vehicle has not completed payment and the order exceeds the preset time limit, the parking management system records the cancellation time of the corresponding order.

[0062] The platform performs regular detection on all parking orders that have not completed payment. When a parking order still fails to complete payment after exceeding the preset time limit following the order generation time, the order is marked as a cancelled state and the cancellation time is recorded. The preset time limit is set according to road parking management rules, and is used to limit the maximum valid duration of the order.

[0063] For example, in a certain urban road parking charging rule, the preset time limit is set to 24 hours. If the order 20XX0701100815 has not completed payment within 24 hours after generation, the order status is updated to the cancelled state at 20XX-07-02 10:08:15, and the cancellation time is recorded as 20XX-07-02 10:08:15.

[0064] The order generation time, payment time or cancellation time, road section identifier, parking space identifier of the parking order, and the corresponding license plate information of the vehicle are associated and encapsulated to generate parking order data for the corresponding vehicle.

[0065] The vehicle license plate information, order generation time, payment time or cancellation time, road section identifier and parking space identifier in the parking order record are associated and encapsulated according to a unified data structure to form a parking order data record. This data structure takes the license plate information as the main identification field, and also includes a time information field and a space identification field, for subsequent data association processing.

[0066] For example, one parking order data record adopts the following structure: {License plate information: "Shanghai AXXXXX", Order generation time: "20XX-07-01 10:08:15", Payment time: "20XX-07-01 10:36:22", Road section identifier: "R-XXXX", Parking space identifier: "P-XXX-03"}.

[0067] Step S4: Associating and integrating the vehicle sensing data, patrol inspection data and parking order data to generate fused positioning information.

[0068] Unify and organize vehicle sensing data, patrol inspection data and parking order data, and establish the correspondence between the three types of data through license plate information and parking space identifiers. The vehicle sensing data provides the time information of the vehicle in the monitoring area, the patrol inspection data provides the time information and position information recorded on site by patrol personnel, and the parking order data provides the parking space identifier and road section identifier information corresponding to the vehicle. By integrating and processing the three types of data, the records of the vehicle in data from different sources form a unified vehicle positioning record.

[0069] For example, for a vehicle with the license plate information "Yue BXXXXX", the appearance time recorded in the vehicle sensing data is 20XX-06-18 09:12:08, the disappearance time is 20XX-06-18 09:37:46, the patrol verification time recorded in the patrol inspection data is 20XX-06-18 09:20:15, the patrol position is longitude 11X.321525 and latitude 2X.124365, and the parking space identifier recorded in the parking order data is P-XXX-05. A complete vehicle parking record can be formed by integrating these data.

[0070] In another embodiment, step S4 can be implemented in the following manner: Using the vehicle license plate information as the primary key, a correlation index is established among the vehicle sensing data, patrol inspection data and parking order data.

[0071] A vehicle data index table is created, the license plate information is used as the index key, and the vehicle sensing data, patrol inspection data and parking order data are grouped and stored according to the license plate information. When the same license plate information exists in data from different sources, the corresponding data records are located through the index table, and these records are merged into an associated data set of the same vehicle.

[0072] For example, the license plate information "Yue BXXXXX" is used as the index key in the index table, and the corresponding vehicle sensing data record, patrol inspection data record and parking order data record are stored in the same data set.

[0073] Using the pre-established parking space number in the parking management system as the spatial anchor point, and based on the preset correspondence between the monitoring area of the roadside terminal and the parking space number, the equipment identifier in the vehicle sensing data, the patrol position in the patrol inspection data, and the parking space identifier in the parking order data are respectively mapped and matched with the corresponding parking space numbers to obtain the matched multi-source data.

[0074] The parking management platform pre-establishes a parking space information table, recording each parking space number, coordinates, and corresponding road segment identifier. It also establishes a mapping table between the monitoring areas of roadside terminals and parking space numbers. Device identifiers retrieve the set of parking space numbers covered by their monitoring area by querying the mapping table; inspection locations are determined by calculating the distance between the inspection location coordinates and the parking space coordinates to find the nearest parking space number; and parking space identifiers in parking order data directly correspond to parking space numbers. Through this method, data from different sources is converted into a unified parking space number representation.

[0075] For example, the monitoring area corresponding to equipment identifier XXX-1 covers berth numbers P-XXX-01 to P-XXX-08; the longitude 11X.321525 and latitude 2X.124365 of the inspection location are matched with berth number P-XXX-05 after distance calculation with berth coordinates; if the parking order data records berth identifier P-XXX-05, then all three types of data correspond to the same berth number P-XXX-05.

[0076] The pre-defined correspondence between the monitoring area of ​​the roadside terminal and the parking space number can be established through on-site calibration during the system deployment phase. Specifically, based on the installation location, shooting or detection range, and actual distribution of road parking spaces of the roadside terminal, the monitoring area of ​​each roadside terminal is spatially marked in the system platform, and the parking space number located within that monitoring area is associated and registered, thereby forming a spatial correspondence between the roadside terminal and the parking space number.

[0077] For example, after installing a roadside terminal on a certain road section, the monitoring coverage area of ​​the roadside terminal can be determined by electronic map or on-site surveying, and the parking space numbers covered by the monitoring area can be marked in the system. When a vehicle is identified by the roadside terminal, the corresponding vehicle perception data can be mapped to the relevant parking space number according to the above correspondence, thereby providing a unified spatial reference for subsequent multi-source data fusion and positioning.

[0078] Based on the order generation time of parking order data, the matched multi-source data is aligned according to the time dimension. The appearance time and disappearance time in vehicle perception data and the inspection verification time in inspection data are incorporated into a unified time axis to obtain time-aligned multi-source data.

[0079] A unified timeline is established, using the order generation time of parking order data as the time reference point. The appearance and disappearance times in vehicle perception data and the inspection verification time in inspection data are converted into time offset values ​​relative to the order generation time. Time records from different data sources are uniformly sorted using these time offset values ​​to form multi-source data records with consistent time.

[0080] For example, the order generation time is 20XX-06-18 09:12:20, the occurrence time of the vehicle sensing data is 09:12:08, the disappearance time is 09:37:46, and the inspection verification time is 09:20:15. By calculating the time difference, the occurrence time offset value can be obtained as -12 seconds, and the inspection verification time offset value as +475 seconds.

[0081] Integrate the matched multi-source data and the time-aligned multi-source data to generate fused positioning information including license plate information, time series information, spatial position information and parking space number.

[0082] Unify and combine the time information in the vehicle sensing data, the spatial position information in the inspection data, and the parking space identification information in the parking order data, and use the parking space number as the spatial identifier to write these fields into the fused positioning data structure, so as to form a complete fused positioning information record.

[0083] For example, a fused positioning information record adopts the following structure: {License plate information: "Yue BXXXXX", Occurrence time: "20XX-06-18 09:12:08", Disappearance time: "20XX-06-18 09:37:46", Inspection verification time: "20XX-06-18 09:20:15", Inspection position: "11X.321525,2X.124365", Parking space number: "P-XXX-05"}.

[0084] Step S5: Correct the fused positioning information through preset time constraints and spatial constraints to obtain corrected fused positioning information.

[0085] Perform rule verification on the time field and spatial field in the fused positioning information, and filter the records in the fused positioning information through time constraints and spatial constraints, so as to eliminate mismatches caused by time recording deviations or spatial positioning deviations between data of different sources. The time constraint is used to verify the consistency of the vehicle sensing data, the inspection data and the parking order data in the time dimension, and the spatial constraint is used to verify the consistency of the inspection position, the roadside terminal monitoring area and the parking space coordinates in the spatial dimension. By performing time verification and spatial verification on the fused positioning information one by one, records that do not meet the constraint conditions are eliminated, and records that meet the constraint conditions are retained and updated, so as to obtain the corrected fused positioning information.

[0086] For example, in a fused positioning information record, the license plate information is "Yue BXXXXX", the appearance time recorded in the vehicle sensing data is 20XX-06-18 09:12:08, the disappearance time is 20XX-06-18 09:37:46, the inspection verification time recorded in the inspection data is 20XX-06-18 09:20:15, the order generation time recorded in the parking order data is 20XX-06-18 09:12:20, the inspection position is at longitude 11X.321525 and latitude 2X.124365, and the parking space number is P-XXX-05. It is judged through the time constraint that the inspection verification time is between the appearance time and the disappearance time, and it is judged through the space constraint that the distance between the inspection position and the parking space coordinate is less than the space threshold, then this record is retained as valid fused positioning information.

[0087] As Figures 3 to 5 shown, in another embodiment, step S5 may be implemented in the following manner: A time range in which the inspection verification time falls within the time interval formed by the appearance time and the disappearance time of the corresponding vehicle sensing data, and the difference between the inspection verification time and the order generation time of the parking order data does not exceed a preset time threshold is taken as the time constraint.

[0088] The time field in the fused positioning information is calculated and verified. The specific calculation manner is: judge whether the inspection verification time satisfies appearance time ≤ inspection verification time ≤ disappearance time, and calculate the time difference between the inspection verification time and the order generation time in the parking order data. The preset time threshold is set according to the time distribution of vehicle parking behaviors in the parking management scenario, and is used to limit the maximum allowable difference between the inspection verification time and the order generation time.

[0089] For example, when the appearance time recorded in the vehicle sensing data is 09:12:08, the disappearance time is 09:37:46, the inspection verification time is 09:20:15, and the order generation time recorded in the parking order data is 09:12:20, it can be obtained through calculation that the inspection verification time is between the appearance time and the disappearance time, and the time difference between the inspection verification time and the order generation time is 475 seconds. If the preset time threshold is set to 900 seconds, this record satisfies the time constraint condition.

[0090] A space range in which the distance between the inspection position in the inspection data and the parking space coordinate corresponding to the parking space identifier in the parking order data does not exceed a preset space threshold, and there is a spatial overlapping relationship between the monitoring area coverage of the roadside terminal and the parking space position corresponding to the parking space identifier is taken as the space constraint.

[0091] Distance calculations and area coverage checks are performed on the spatial fields in the fused positioning information. Berth coordinates are obtained from a berth information table, which records the longitude and latitude coordinates corresponding to each berth number. Inspection locations are obtained through the satellite positioning module of the inspection terminal. Whether the spatial threshold condition is met is determined by calculating the geographic distance between the inspection location and the berth coordinates. The distance calculation uses the Haversine distance calculation algorithm to calculate the spherical distance between two geographic coordinate points based on longitude and latitude. The preset spatial threshold is set based on the road berth width and the inspection positioning error. Monitoring area coverage is determined by querying the device monitoring area table, which records the boundary coordinates of the monitoring area corresponding to each device identifier.

[0092] For example, the inspection location is at longitude 11X.321525, latitude 2X.124365, and the berth coordinates are at longitude 11X.321510, latitude 2X.124350. The distance between the two points, calculated using Haversine distance, is 2.3 meters. When the preset spatial threshold is set to 5 meters, this distance satisfies the spatial constraint condition. Furthermore, a query of the equipment monitoring area table shows that the monitoring area of ​​equipment identifier XXX-1 covers the area of ​​berth number P-XXX-05; therefore, the spatial constraint condition is deemed valid.

[0093] Each record in the fused positioning information is verified sequentially using time and space constraints. Records that simultaneously meet both time and space constraints are retained, and the retained records are used to update the corrected fused positioning information.

[0094] Perform rule validation on each record in the fused positioning information data table. Calculate the temporal and spatial constraints for each record sequentially. If both constraints are met, mark the record as valid and write it into the corrected fused positioning information data table. If either constraint is not met, do not include the record in the corrected fused positioning information data table.

[0095] For example, in the fused positioning information data table, there are three records: record 1 satisfies both time and spatial constraints; record 2 satisfies only the time constraint but not the spatial constraint; and record 3 satisfies the spatial constraint, but the inspection and verification time exceeds the time threshold. After verification, only record 1 is retained and written into the corrected fused positioning information data table.

[0096] Records that do not meet any constraint rule are marked as abnormal and a manual review process is triggered.

[0097] Set an abnormality flag for records that fail constraint verification, and write the abnormal records into an abnormal record table. The abnormal record table stores fields including license plate information, inspection verification time, inspection position, parking space number, and abnormality cause. The abnormality cause field is used to identify the violated constraint type, such as time constraint abnormality or space constraint abnormality. Manual reviewers check the data through the abnormal record table, and correct or confirm the records.

[0098] For example, when the distance between the inspection position and the parking space coordinate in a certain fused positioning information record is 18 meters, which is greater than the preset space threshold of 5 meters, the record is marked as a space constraint abnormality and written into the abnormal record table, and the abnormality cause field is recorded as "space constraint abnormality".

[0099] Step S6: Map the corrected fused positioning information to a preset digital road network structure model to determine the actual parking space position of the vehicle in the digital road network.

[0100] Perform spatial positioning on the parking space number in the corrected fused positioning information through the digital road network structure model, and obtain the spatial position of the corresponding parking space and the affiliated road information according to the parking space number. By writing the correspondence between the license plate information and the parking space number into the parking space node in the digital road network structure model, the parking position of the vehicle in the road network can be accurately identified in the digital road network structure model.

[0101] For example, there is a record in the corrected fused positioning information: the license plate information is "Yue BXXXXX", the parking space number is P-XXX-05, and the parking space coordinate corresponding to the parking space number can be obtained by querying the digital road network structure model as longitude 11X.321510, latitude 2X.124350, and the affiliated road section identifier is R-XXX, so as to determine the parking position of the vehicle in the digital road network.

[0102] As Figure 6 shown, in another embodiment, step S6 can be implemented in the following manner: A digital road network structure model is pre-constructed. The digital road network structure model takes road sections as basic units, assigns a unique road section identifier to each road section, and registers the parking space information and parking space number of each parking space in sequence on each road section according to the actual parking space distribution, establishes a hierarchical spatial correspondence between the road section identifier and the parking space number, and forms a three-level index structure of road network-road section-parking space.

[0103] A digital road network structure model is constructed based on road basic data. The road basic data includes road centerline data, road boundary data and road parking space layout data. A road section identifier is assigned to each road according to the road management number, and the parking space number, parking space coordinate and parking space direction information are recorded on each road section according to the parking space layout diagram. By establishing the mapping relationship between the road section identifier and the parking space number, each parking space number can be located to its affiliated road section.

[0104] For example, parking space numbers P-XXX-01 to P-XXX-08 are registered under the road number R-XXX, and each parking space number corresponds to a set of parking space coordinates. For example, the parking space number P-XXX-05 corresponds to the parking space coordinates with a longitude of 11X.321510 and a latitude of 2X.124350.

[0105] Taking the parking space number in the corrected fused positioning information as a search key, a search is performed in the preset three-level index structure of the digital road network structure model based on the search key, so as to obtain the road section identifier and parking space information corresponding to the parking space number.

[0106] A search is performed in the parking space information table through the parking space number, and the road section identifier, parking space coordinate and parking space serial number information corresponding to the parking space number are obtained. The parking space information table is a basic data table in the digital road network structure model, which is used to record the corresponding relationship between the parking space number and the road section identifier.

[0107] For example, when the parking space number is P-XXX-05, the affiliated road section identifier obtained by querying the parking space information table is R-XXX, and the parking space coordinates are a longitude of 11X.321510 and a latitude of 2X.124350.

[0108] The license plate information in the corrected fused positioning information is bound with the road section identifier and parking space information obtained by search, so as to generate an actual parked parking space position, wherein the actual parked parking space position includes the license plate information, the affiliated road section identifier, the specific parking space number and the parking space coordinate.

[0109] The license plate information field and the parking space number field are combined, and the road section identifier and the parking space coordinate information are appended, so as to form a vehicle parking position record. This record is used to describe the specific parking position of the vehicle in the road network.

[0110] For example, one actual parked parking space position record is as follows: {license plate information: "Yue BXXXXX", road section identifier: "R-XXX", parking space number: "P-XXX-05", parking space coordinate: "11X.321510, 2X.124350"}.

[0111] The actual parked parking space position is input into the corresponding node of the digital road network structure model, so as to realize the visual expression and dynamic update of the actual parked parking space position of the vehicle in the digital road network.

[0112] By writing the recorded actual parking berth positions into the berth node data table in the digital road network structure model, an association relationship is established between the berth nodes and the corresponding vehicle license plate information. On the digital map interface, vehicle identification icons are drawn according to the coordinate positions of the berth nodes, and real-time update of vehicle parking positions is realized by refreshing the berth node data table.

[0113] For example, in the digital map interface, when the berth node with berth number P-XXX-05 is associated with the license plate information "Yue BXXXXX", a vehicle icon is displayed at the coordinate position of this berth, and the vehicle identification is removed when the vehicle leaves the berth or the order ends, so as to complete the dynamic update in the digital road network.

[0114] In this embodiment, roadside terminals continuously collect vehicle license plate information and record the appearance time and disappearance time of the corresponding vehicles to construct vehicle sensing data. At the same time, the license plate information, inspection verification time and inspection position reported by inspection personnel are received to form inspection data, and the license plate information of vehicles in parking orders, order generation time, payment time or cancellation time, as well as the road section identifier and berth identifier associated with the parking order are acquired to construct parking order data. On this basis, the vehicle sensing data, inspection data and parking order data are uniformly associated and integrated to form fused positioning information including vehicle identity information, time information and berth information. Subsequently, consistency verification and correction are performed on the fused positioning information through preset time constraints and space constraints, and abnormal records that do not meet the constraint conditions are eliminated, so as to obtain reliable corrected fused positioning information. Further, the corrected fused positioning information is mapped to the preset digital road network structure model, and the vehicle parking position is positioned and bound through the hierarchical index relationship of road network - road section - berth, so as to finally determine the actual parking berth position of the vehicle in the digital road network, and realize the visual expression and dynamic update of the vehicle parking position in the digital road network. Through the above method, effective fusion and mutual verification among vehicle sensing data, inspection data and parking order data are realized, which can improve the accuracy of vehicle parking position identification and data consistency in complex roadside parking scenarios, and provide a reliable data basis for digital road network management and parking supervision.

[0115] As Figure 7 shown, in another example, the present application further provides a vehicle positioning system 10 for road network holographic sensing, comprising: Roadside collection module 11, configured to collect license plate information of vehicles through roadside terminals, and record the appearance time and disappearance time of the corresponding vehicles to construct vehicle sensing data.

[0116] Personnel collection module 12, configured to receive the license plate information of vehicles, inspection verification time and inspection position reported by inspection personnel to construct inspection data.

[0117] The order collection module 13 is used to obtain the license plate information of the vehicle in the parking order, the order generation time, the payment time or cancellation time, as well as the road segment identifier and parking space identifier associated with the parking order, in order to construct parking order data.

[0118] The data integration module 14 is used to link and integrate vehicle perception data, inspection data and parking order data to generate fused positioning information.

[0119] The constraint correction module 15 is used to correct the fused positioning information by means of preset time constraints and spatial constraints, so as to obtain the corrected fused positioning information.

[0120] The data mapping module 16 is used to map the corrected fused positioning information to a preset digital road network structure model in order to determine the actual parking space location of the vehicle in the digital road network.

[0121] The roadside data acquisition module 11, personnel data acquisition module 12, and order data acquisition module 13 respectively acquire vehicle perception data, inspection data, and parking order data from roadside terminals, inspection personnel, and parking management systems, thus forming multi-source parking information from different sources that can complement each other. Subsequently, the data integration module 14 uniformly associates and integrates the vehicle perception data, inspection data, and parking order data, so that the data from different sources form corresponding relationships at the vehicle identity, time information, and spatial information levels, and generates fused positioning information. On this basis, the constraint correction module 15 uses preset time constraints and spatial constraints to perform consistency verification and correction on the fused positioning information, and filters out data records that simultaneously satisfy the time logic relationship and spatial correspondence relationship, thereby obtaining reliable corrected fused positioning information. Finally, the data mapping module 16 maps the corrected fused positioning information to a preset digital road network structure model, and determines the actual parking space location of the vehicle in the digital road network according to the hierarchical relationship between road segment identifiers and parking space numbers in the digital road network structure model. Through the collaborative data processing among the above system modules, the effective integration and mutual verification of vehicle perception data, inspection data, and parking order data are realized, thereby improving the accuracy of vehicle parking location identification and data consistency, and providing reliable data support for digital road network management and parking resource supervision.

[0122] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims can be performed in a different order than that shown in the embodiments and still achieve the desired results. Additionally, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are also possible or may be advantageous.

[0123] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0124] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware, or by a program instructing the relevant hardware to implement them. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0125] The above are merely preferred embodiments of this application and are not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A vehicle localization method based on road network holographic perception, characterized in that, include: The roadside terminal uses a built-in image acquisition module to continuously acquire road images within its monitoring area, obtaining a continuous frame image sequence. Vehicle target detection is performed on each frame of the continuous frame image sequence to extract candidate image blocks of the license plate area, and character recognition is performed on the candidate image blocks to obtain the license plate information of the corresponding vehicle. The system timestamp corresponding to the first recognition of the license plate information is recorded as the appearance time of the vehicle; the continuous frame image sequence is continuously tracked and detected, and when no vehicle corresponding to the license plate information is detected within a preset number of consecutive frames, the system timestamp corresponding to the last successful recognition of the license plate information is recorded as the disappearance time of the vehicle; The license plate information, the appearance time, the disappearance time, and the device identifier of the roadside terminal are associated and encapsulated to generate vehicle perception data for the corresponding vehicle. Receive vehicle license plate information, inspection verification time, and inspection location reported by inspection personnel to construct inspection data; Obtain the license plate information of the vehicle in the parking order, the order generation time, the payment time or cancellation time, as well as the road segment identifier and parking space identifier associated with the parking order, to construct parking order data; Using pre-established parking space numbers as spatial anchors, and based on the preset correspondence between the monitoring area of ​​the roadside terminal and the parking space numbers, the device identifier in the vehicle perception data, the inspection location in the inspection data, and the parking space identifier in the parking order data are mapped and matched with the corresponding parking space numbers to obtain matched multi-source data. Taking the order generation time of the parking order data as the benchmark, the matched multi-source data is aligned according to the time dimension, and the appearance time and disappearance time in the vehicle perception data and the inspection verification time in the inspection data are incorporated into a unified time axis to obtain time-aligned multi-source data. The matched multi-source data and the time-aligned multi-source data are integrated to generate fused positioning information; The time constraint is defined as follows: the inspection verification time falls within the time interval formed by the appearance time and disappearance time of the corresponding vehicle perception data, and the difference between the inspection verification time and the order generation time of the parking order data does not exceed a preset time threshold. The spatial constraint is defined as the spatial range in which the distance between the inspection location in the inspection data and the parking space coordinates corresponding to the parking space identifier in the parking order data does not exceed a preset spatial threshold, and the spatial range in which the monitoring area coverage of the roadside terminal overlaps with the parking space location corresponding to the parking space identifier. Each record in the fused positioning information is verified sequentially using the time constraint and the spatial constraint. Records that simultaneously meet both the time constraint and the spatial constraint are retained, and the retained records are updated as the corrected fused positioning information. Records that do not meet either constraint rule are marked as abnormal and a manual review process is triggered. Using the parking space number in the corrected fused positioning information as the search key, a search is performed in a preset digital road network structure model based on the search key to obtain the road segment identifier and parking space information corresponding to the parking space number; the license plate information in the corrected fused positioning information is bound with the retrieved road segment identifier and parking space information to generate the actual parking space location, wherein the actual parking space location includes license plate information, the road segment identifier, the specific parking space number, and the parking space coordinates; the actual parking space location is input into the corresponding node of the digital road network structure model to realize the visual representation of the actual parking space location of the vehicle in the digital road network.

2. The vehicle positioning method based on road network holographic perception according to claim 1, characterized in that, The step of receiving vehicle license plate information, inspection verification time, and inspection location reported by inspection personnel to construct inspection data includes: Patrol personnel use patrol terminals to conduct on-site inspections of parked vehicles along a preset patrol route. The image acquisition module built into the patrol terminal is used to capture images of the license plate area of ​​the parked vehicles, and the captured images are used for character recognition to obtain the license plate information of the corresponding vehicles. The system timestamp is obtained as the inspection and verification time, and the positioning module built into the inspection terminal is called to obtain the current geographic coordinates as the inspection location. The license plate information, the inspection and verification time, the inspection location, and the employee ID of the inspection personnel are associated and encapsulated to form the inspection data for the corresponding vehicle.

3. The vehicle positioning method based on road network holographic perception according to claim 1, characterized in that, The step of obtaining the vehicle's license plate information, order generation time, payment time or cancellation time, and the road segment identifier and parking space identifier associated with the parking order to construct parking order data includes: When a vehicle enters a parking space, the corresponding parking space sensing device triggers an order generation event, records the order generation time, the parking space identifier bound to the parking space, and the road segment identifier of the road segment to which the parking space belongs, and forms a parking order. When the vehicle leaves after payment, record the payment time for the corresponding order; If a vehicle fails to complete payment and the order exceeds the preset time limit, record the cancellation time of the corresponding order; The parking order data is generated by associating and encapsulating the order generation time, payment time or cancellation time, road segment identifier, parking space identifier, and the license plate information of the vehicle.

4. A vehicle positioning system based on road network holographic perception, characterized in that, include: The roadside acquisition module is used to continuously acquire road images within its monitoring area through the built-in image acquisition module of the roadside terminal, and obtain a continuous frame image sequence. Vehicle target detection is performed on each frame of the continuous frame image sequence to extract candidate image blocks of the license plate area, and character recognition is performed on the candidate image blocks to obtain the license plate information of the corresponding vehicle. The system timestamp corresponding to the first recognition of the license plate information is recorded as the appearance time of the vehicle; the continuous frame image sequence is continuously tracked and detected, and when no vehicle corresponding to the license plate information is detected within a preset number of consecutive frames, the system timestamp corresponding to the last successful recognition of the license plate information is recorded as the disappearance time of the vehicle; The license plate information, the appearance time, the disappearance time, and the device identifier of the roadside terminal are associated and encapsulated to generate vehicle perception data for the corresponding vehicle. The personnel data collection module is used to receive vehicle license plate information, inspection verification time, and inspection location reported by inspection personnel in order to construct inspection data. The order acquisition module is used to obtain the license plate information of the vehicle in the parking order, the order generation time, the payment time or cancellation time, as well as the road segment identifier and parking space identifier associated with the parking order, in order to construct parking order data; The data integration module is used to map and match the device identifier in the vehicle perception data, the inspection location in the inspection data, and the parking space identifier in the parking order data with the corresponding parking space number, based on the pre-established parking space number as a spatial anchor point and the preset correspondence between the monitoring area of ​​the roadside terminal and the parking space number, to obtain matched multi-source data. Using the order generation time of the parking order data as a benchmark, the matched multi-source data is aligned according to the time dimension, incorporating the appearance time and disappearance time in the vehicle perception data and the inspection verification time in the inspection data into a unified time axis to obtain time-aligned multi-source data. The matched multi-source data and the time-aligned multi-source data are integrated to generate fused positioning information; The constraint correction module is used to define a time constraint where the inspection verification time falls within the time interval formed by the appearance and disappearance times of the corresponding vehicle perception data, and the difference between the inspection verification time and the order generation time of the parking order data does not exceed a preset time threshold; a spatial constraint where the distance between the inspection location in the inspection data and the parking space coordinates corresponding to the parking space identifier in the parking order data does not exceed a preset spatial threshold, and the monitoring area coverage of the roadside terminal overlaps with the parking space location corresponding to the parking space identifier; each record in the fused positioning information is sequentially verified using the time and spatial constraints, and records that simultaneously meet both the time and spatial constraints are retained, and the retained records are updated as the corrected fused positioning information; records that do not meet any constraint rule are marked as abnormal and a manual review process is triggered; The data mapping module is used to retrieve the road segment identifier and parking space information corresponding to the parking space number in the corrected fused positioning information using the parking space number as the search key in a preset digital road network structure model. It then binds the license plate information in the corrected fused positioning information with the retrieved road segment identifier and parking space information to generate the actual parking space location. The actual parking space location includes the license plate information, the road segment identifier, the specific parking space number, and the parking space coordinates. Finally, it inputs the actual parking space location into the corresponding node of the digital road network structure model to achieve a visual representation of the vehicle's actual parking space location in the digital road network.

Images