CN114663862A - Vehicle identification method, device, system and storage medium - Google Patents

Abstract

A vehicle identification method, a device, a storage medium and a system are provided, and the vehicle identification method comprises the following steps: when a unlicensed vehicle enters a field, acquiring an image of the vehicle to generate first characteristic information and wireless fidelity media access control information to generate a first wireless fidelity media access control information set which is used as a field entry record and stored, and when the vehicle is in the field, generating the field record and storing; when the unlicensed vehicle leaves, acquiring second characteristic information and a second wireless fidelity medium access control information set, and when more than one entry record is matched, acquiring a second characteristic information sequence of the current vehicle and the vehicles which have left the vehicle; searching a similar third characteristic information sequence, and determining a wireless fidelity media access control information total set according to the third characteristic information sequence; and charging according to the entrance record matched with the total set. According to the scheme provided by the embodiment, the vehicle is identified through the characteristic information and the WiFi MAC, manual participation is not needed, and the passing efficiency is high.

Description

Vehicle identification method, device, system and storage medium

Technical Field

The present disclosure relates to parking charging technologies, and in particular, to a method, an apparatus, a system, and a storage medium for vehicle identification.

Background

In order to improve the vehicle passing efficiency, license plate automatic identification systems are usually installed at entrances and exits of current large-scale business supermarkets, residences and park parks. According to the license plate information and the parking duration, the system can generate corresponding parking cost. However, in reality, the situation that the license plate is stained or not hung always occurs, and at this time, the driver and the passenger generally need to get tickets/register and enter and exit in a mode that the two-dimensional code is scanned by a mobile phone because the license plate information cannot be acquired. Because the link of self-service operation of personnel is added, the process is complex and uncontrollable, and the passing efficiency and the user experience are influenced.

Another solution is to associate the facial image of the driver with the vehicle type information and start calculating the parking time and the fee by collecting the facial image and the vehicle type information of the driver of the vehicle. The scheme relates to the collection of biological information and has legal risks; in addition, in order to ensure the automatic acquisition effect of the human face in the automobile, a flashing light needs to be configured in the system, which causes light pollution to drivers and passengers; and if the system needs to actively cooperate the entrance and the exit of the driver (such as a windowing probe or a getting-off car), the blind passing experience of the entrance and the exit can be influenced.

Disclosure of Invention

The embodiment of the application provides a vehicle identification method, a device, a system and a storage medium, which can improve the passing efficiency of unlicensed vehicles.

The embodiment of the application provides a vehicle identification method, which comprises the following steps:

when a unlicensed vehicle enters, acquiring an image of the vehicle to generate first characteristic information, acquiring wireless fidelity media access control information of the vehicle and the periphery to generate a first wireless fidelity media access control information set, and generating and storing an entry record of the vehicle, wherein the entry record comprises the first characteristic information, the first wireless fidelity media access control information set and the entry time of the vehicle; when a vehicle is in a field, generating third characteristic information according to an image of the vehicle acquired by a camera through which the vehicle passes, acquiring wireless fidelity media access control information of the vehicle and the periphery of the vehicle to generate a third wireless fidelity media access control information set, and generating and storing an in-field record of the vehicle, wherein the in-field record comprises identification information of the camera, the third characteristic information, the third wireless fidelity media access control information set and an image timestamp of the image of the vehicle acquired by the camera;

when the unlicensed vehicle leaves, acquiring an image of the current vehicle leaving to generate second characteristic information, and acquiring the current vehicle leaving and the wireless fidelity media access control information of the periphery to generate a second wireless fidelity media access control information set; matching the second characteristic information with entry records, when the second characteristic information is matched with a plurality of entry records and the second wireless fidelity media access control information set is intersected with the first wireless fidelity media access control information set of more than one entry record in the matched entry records, acquiring the second characteristic information of at least one vehicle which is out of the field before the vehicle which is out of the field currently, generating a second characteristic information sequence by the vehicle which is out of the field currently and the second characteristic information of the vehicle which is out of the field currently according to the field-out sequence, wherein the vehicles corresponding to the second characteristic information in the second characteristic information sequence are a plurality of continuous vehicles which are out of the field sequentially;

searching vehicles passing through the cameras for a plurality of cameras preset in a field, and merging a wireless fidelity media access control information set corresponding to each piece of third feature information in a third feature information sequence according to the in-field record of a target camera to generate a wireless fidelity media access control information total set when the target camera exists and the target camera has the third feature information sequence formed by the third feature information of the passing vehicles according to the passing sequence and the ith third feature information in the third feature information sequence is similar to the ith second feature information in the second feature information sequence; and when the total wireless fidelity media access control information set only has intersection with a first wireless fidelity media access control information set of one of the matched entry records, charging the vehicles on departure according to the entry time of the entry record with the intersection, wherein the passing sequence is consistent with the time sequence of the departure sequence, i is 1 to n, n is the number of second characteristic information in the second characteristic information sequence, and n is greater than 1.

In an exemplary embodiment, the third characteristic information sequence is a sequence of third characteristic information of vehicles that continuously approach the target camera.

In an exemplary embodiment, the finding the vehicle passing through the camera includes: and searching for vehicles passing through the camera within a preset time period.

In an exemplary embodiment, the predetermined plurality of cameras in the field includes a plurality of cameras near the field.

In an exemplary embodiment, the method further comprises: and when the second characteristic information is matched with a plurality of entrance records and the second wireless fidelity media access control information set only has intersection with the first wireless fidelity media access control information set of one entrance record in the matched plurality of entrance records, charging the vehicles which come out of the vehicle according to the entrance time of the entrance record with the intersection.

In an exemplary embodiment, the method further comprises: and when the second characteristic information is only matched with one entry record, charging the vehicles for leaving according to the entry time of the matched entry record.

In an exemplary embodiment, the first feature information, the second feature information, and the third feature information are semi-structured feature vectors generated by a preset neural network model.

The embodiment of the present disclosure provides a vehicle identification device, which includes a memory and a processor, where the memory stores a program, and the program, when read and executed by the processor, implements the vehicle identification method according to any one of the embodiments.

Embodiments of the present disclosure provide a computer-readable storage medium storing one or more programs, which are executable by one or more processors to implement a vehicle identification method as in any of the above embodiments.

The disclosed embodiment provides a vehicle identification system, including a camera at an entrance, a camera at an exit, at least one camera in a field, and a background server, wherein:

the camera at the entrance is configured to acquire an image of the vehicle when an unlicensed vehicle enters the entrance, or acquire the image of the vehicle to generate first characteristic information, and send the image or the first characteristic information and the entrance time of the vehicle to the background server; acquiring wireless fidelity media access control information of the vehicle and the periphery, generating a first wireless fidelity media access control information set, and sending the first wireless fidelity media access control information set to the background server;

the camera at the exit is configured to acquire an image of a vehicle which is currently present when an unlicensed vehicle is present, or acquire second characteristic information generated by the vehicle which is currently present, send the image or the second characteristic information to the background server, acquire wireless fidelity media access control information of the vehicle which is currently present and the periphery, generate a second wireless fidelity media access control information set and send the second wireless fidelity media access control information set to the background server;

the cameras in the field are configured to acquire images of passing vehicles when the unlicensed vehicles pass by, or acquire the images of the vehicles to generate third feature information, and send the images or the third feature information and image timestamps of the images acquired by the cameras to the background server; acquiring the passing vehicles and the surrounding wireless fidelity media access control information to generate a third wireless fidelity media access control information set and sending the third wireless fidelity media access control information set to the background server;

the background server is configured to receive an image sent by a camera at an entrance to generate first characteristic information, or receive the first characteristic information sent by the camera at the entrance, receive a first wireless fidelity medium access control information set sent by the camera at the entrance, generate and store an entry record of a vehicle, wherein the entry record comprises the first characteristic information, the first wireless fidelity medium access control information set and the entry time of the vehicle;

receiving an image sent by a camera in a field to generate third characteristic information or receiving third characteristic information sent by the camera in the field, receiving a third wireless fidelity medium access control information set sent by the camera in the field, generating and storing a field record of the vehicle, wherein the field record comprises identification information of the camera in the field, the third characteristic information and the third wireless fidelity medium access control information set, and the camera acquires an image time stamp of the image of the passing vehicle;

receiving an image sent by the camera at the exit to generate second characteristic information or receiving the second characteristic information sent by the camera at the exit, and receiving a second wireless fidelity medium access control information set sent by the camera at the exit;

when the second characteristic information is matched with a plurality of entrance records and the intersection exists between the second wireless fidelity media access control information set and a first wireless fidelity media access control information set of more than one entrance record in the plurality of matched entrance records, acquiring second characteristic information of at least one vehicle which is out of the field before the vehicle which is out of the field currently, and generating a second characteristic information sequence according to the appearance sequence by using the second characteristic information of the vehicle which is out of the field currently and the second characteristic information of the vehicle which is out of the field already; the vehicles corresponding to the second characteristic information in the second characteristic information sequence are a plurality of continuous vehicles which come out of the field in sequence;

searching vehicles passing through the cameras for a plurality of cameras preset in a field, and merging a wireless fidelity media access control information set corresponding to each piece of third feature information in a third feature information sequence according to the in-field record of a target camera to generate a wireless fidelity media access control information total set when the target camera exists and the target camera has the third feature information sequence formed by the third feature information of the passing vehicles according to the passing sequence and the ith third feature information in the third feature information sequence is similar to the ith second feature information in the second feature information sequence; and when the total wireless fidelity media access control information set only has intersection with a first wireless fidelity media access control information set of one of the matched entry records, charging the vehicles on departure according to the entry time of the entry record with the intersection, wherein the passing sequence is consistent with the time sequence of the departure sequence, i is 1 to n, n is the number of second characteristic information in the second characteristic information sequence, and n is greater than 1.

Compared with the related art, the vehicle identification method and device, the storage medium and the system are provided, and the vehicle identification method comprises the following steps: when a unlicensed vehicle enters a field, acquiring an image of the vehicle to generate first characteristic information, acquiring wireless fidelity media access control information of the vehicle and the periphery to generate a first wireless fidelity media access control information set, generating and storing an entrance record of the vehicle, and when the vehicle is in the field, generating and storing an in-field record of the vehicle; when the unlicensed vehicle leaves, acquiring an image of the vehicle which leaves at present to generate second characteristic information, and acquiring the vehicle which leaves at present and wireless fidelity media access control information of the periphery to generate a second wireless fidelity media access control information set; matching the second characteristic information with entry records, when the second characteristic information is matched with a plurality of entry records and the second wireless fidelity media access control information set is intersected with the first wireless fidelity media access control information set of more than one entry record in the matched entry records, acquiring the second characteristic information of at least one vehicle which is out of the field before the vehicle which is out of the field currently, and generating a second characteristic information sequence by the vehicle which is out of the field currently and the second characteristic information of the vehicle which is out of the field currently according to the field-out sequence; searching a similar third characteristic information sequence, and determining a wireless fidelity media access control information total set according to the third characteristic information sequence; and charging according to the intersection of the total wireless fidelity media access control information set and the entrance record. The scheme that this embodiment provided carries out the discernment of vehicle through characteristic information and wiFi MAC, need not artifical the participation (manual registration, sweep the sign indicating number etc.), and the current efficiency is high, and user experience is good, can discern the unlicensed vehicle of the same style, and need not discernment user's biological information, can avoid the law risk.

Additional features and advantages of the application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the application. Other advantages of the present application may be realized and attained by the instrumentalities and combinations particularly pointed out in the specification and the drawings.

Drawings

The accompanying drawings are included to provide an understanding of the present disclosure and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the examples serve to explain the principles of the disclosure and not to limit the disclosure.

FIG. 1 is a flow chart of a method for vehicle identification provided in an exemplary embodiment;

FIG. 2 is a schematic illustration of a vehicle identification system provided in an exemplary embodiment;

FIG. 3 is a block diagram of a vehicle identification device according to an exemplary embodiment.

Detailed Description

The description herein describes embodiments, but is intended to be exemplary, rather than limiting and it will be apparent to those of ordinary skill in the art that many more embodiments and implementations are possible that are within the scope of the embodiments described herein. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are possible. Any feature or element of any embodiment may be used in combination with, or instead of, any other feature or element in any other embodiment, unless expressly limited otherwise.

The present application includes and contemplates combinations of features and elements known to those of ordinary skill in the art. The embodiments, features and elements disclosed in this application may also be combined with any conventional features or elements to form a unique inventive concept as defined by the claims. Any feature or element of any embodiment may also be combined with features or elements from other inventive aspects to form yet another unique inventive aspect, as defined by the claims. Thus, it should be understood that any of the features shown and/or discussed in this application may be implemented alone or in any suitable combination. Accordingly, the embodiments are not limited except as by the appended claims and their equivalents. Furthermore, various modifications and changes may be made within the scope of the appended claims.

Further, in describing representative embodiments, the specification may have presented the method and/or process as a particular sequence of steps. However, to the extent that the method or process does not rely on the particular order of steps set forth herein, the method or process should not be limited to the particular sequence of steps described. Other sequences of steps are possible as will be appreciated by those of ordinary skill in the art. Therefore, the particular order of the steps set forth in the specification should not be construed as limitations on the claims. Furthermore, the claims directed to the method and/or process should not be limited to the performance of their steps in the order written, and one skilled in the art can readily appreciate that the sequences may be varied and still remain within the spirit and scope of the embodiments of the present application.

The embodiment of the disclosure provides a vehicle identification implementation method, which is based on vehicle identification and WiFi wireless sniffing to cooperatively identify vehicles, can realize the same passing efficiency as a branded vehicle, and avoids legal risks possibly faced by a face acquisition scheme.

Fig. 1 is a flowchart of a vehicle identification method according to an embodiment of the present disclosure. As shown in fig. 1, the method for identifying a unlicensed vehicle and parking fee provided by this embodiment includes:

step 101, when an unlicensed vehicle enters, acquiring an image of the vehicle to generate first feature information, acquiring WiFi MAC information of the vehicle and the periphery of the vehicle to generate a first WiFi MAC information set, generating and storing an entry record of the vehicle, wherein the entry record comprises the first feature information, the first WiFi MAC information set and entry time of the vehicle;

step 102, when a vehicle is in a field, generating third characteristic information according to an image of the vehicle acquired by a camera through which the vehicle passes, generating a third WiFi MAC information set by the vehicle and WiFi MACs around the vehicle, generating and storing an in-field record of the vehicle, wherein the in-field record comprises identification information of the camera, the third characteristic information and the third WiFi MAC information set, and the camera acquires an image timestamp of the image of the vehicle;

103, when the unlicensed vehicle leaves, acquiring an image of the vehicle which is currently leaving to generate second feature information, acquiring the vehicle which is currently leaving and WiFi MACs around the vehicle to generate a second WiFi MAC information set, matching the second feature information with entry records, acquiring second feature information of at least one vehicle which is already leaving before the vehicle which is currently leaving when the second feature information is matched with a plurality of entry records and the second WiFi MAC information set is intersected with a first WiFi MAC information set of more than one entry record in the plurality of matched entry records, and generating a second feature information sequence according to a leaving sequence by using the second feature information of the vehicle which is currently leaving and the second feature information of the vehicle which is already leaving; the vehicles corresponding to the second characteristic information in the second characteristic information sequence are a plurality of continuous vehicles which come out of the field in sequence;

104, searching vehicles passing through the cameras for a plurality of cameras preset in a field, and merging WiFi MAC information sets corresponding to each piece of third feature information in a third feature information sequence according to field records of the target cameras to generate a WiFi MAC total set when one target camera exists, the target camera has the third feature information sequence formed by the third feature information of the passing vehicles according to the passing sequence, and the ith third feature information in the third feature information sequence is similar to the ith second feature information in the second feature information sequence; and when the WiFi MAC total set only has intersection with the first WiFi MAC information set of one entry record in the matched entry records, charging the vehicles which come out according to the entry time of the entry record with the intersection. Wherein i is 1 to n, n is the number of the second feature information in the second feature information sequence, n is greater than 1, and the passing sequence is consistent with the time sequence of the departure sequence. That is, when the departure sequence is arranged in sequence from the earliest departure to the latest departure, the route sequence is arranged in sequence from the earliest route to the latest route; and vice versa.

The scheme that this embodiment provided carries out the discernment of vehicle through characteristic information and wiFi MAC, need not artifical the participation (manual registration, sweep the sign indicating number etc.), and the current efficiency is high, and user experience is good, can discern the unlicensed vehicle of the same style, and need not discernment user's biological information, can avoid the law risk.

In the embodiment, the field is a parking lot, and cameras are arranged at the entrance and the exit of the parking lot for image acquisition; one or more cameras may be located within the parking lot.

In an exemplary embodiment, the unlicensed vehicle may be detected by license plate location and identification. When the license plate can not be identified, the vehicle is a license-free vehicle.

In an exemplary embodiment, the method further comprises, when a vehicle with a license plate enters the field, identifying the license plate of the vehicle, and recording the license plate and the time of the entrance of the vehicle; when a vehicle with a license plate leaves, the license plate of the vehicle is identified, and charging is carried out according to the entrance time corresponding to the license plate. The scheme provided by the embodiment can charge the vehicles with the license plates directly according to the license plates, and is simple and convenient to realize and high in efficiency. But the embodiment of the present disclosure is not limited thereto, and a licensed vehicle may be identified according to the characteristic information and the WIFI MAC information.

In an exemplary embodiment, the third characteristic information sequence is a sequence of third characteristic information of vehicles that continuously approach the target camera. However, the embodiment of the present disclosure is not limited thereto, for example, the third feature information in the third feature information sequence corresponds to vehicles 1,2, and 3, and vehicles 1,2, and 3 may be vehicles continuously passing through the target camera, or vehicles 4 passing through the target camera between vehicles 1 and 2.

In an exemplary embodiment, when the unlicensed vehicle leaves, acquiring the image of the vehicle that is currently leaving to generate the second feature information, and acquiring the WiFi MACs of the vehicle that is currently leaving to and the surroundings to generate the second WiFi MAC information set includes:

acquiring an image of a current vehicle on the spot to generate second characteristic information, and acquiring the current vehicle on the spot and WiFi MAC information around the current vehicle on the spot to generate a second WiFi MAC information set when the second characteristic information is matched with a plurality of entrance records;

the method further comprises the following steps: and when the second characteristic information is only matched with one entry record, charging the vehicles for leaving according to the entry time of the matched entry record.

The entry record matched with the second characteristic information is as follows: and the first characteristic information and the second characteristic information to be matched are similar to each other, wherein the similarity of the characteristic information can be that the two pieces of characteristic information meet a preset matching condition. Taking the feature information as an example of the feature vector, when comparing the feature vectors, if the euclidean distance or the cosine distance between the two is less than or equal to a preset threshold, the two feature vectors are similar. The similarity between the two pieces of feature information includes a case where the two pieces of feature information are the same.

In this embodiment, when only one entry record is matched according to the second feature information, the identification and the charging can be performed without acquiring WiFi MAC information, so that the charging can be simplified.

In an exemplary embodiment, the matching the second feature information with the entry record includes: and calculating the Euclidean distance or cosine distance between the second characteristic information and the first characteristic information of the entry record, wherein when the Euclidean distance or cosine distance is less than or equal to a preset threshold value, the second characteristic information is matched with the entry record.

In an exemplary embodiment, the first feature information, the second feature information, and the third feature information may be semi-structured feature vectors generated by a preset neural network model. The preset neural network model can be a neural network model obtained through training, and the semi-structured feature vector can be obtained by inputting the obtained image into the preset neural network model. However, the embodiment of the present disclosure is not limited thereto, and the first feature information and the second feature information may be vehicle type information obtained based on neural network model recognition. The semi-structured feature vector is used, and in the face of new vehicle styles, manufacturers do not need to introduce new vehicle data to supervise and learn the neural network model any more, the trouble of periodically upgrading parking lot equipment is avoided, and compared with the semi-structured feature vector obtained through images, the semi-structured feature vector obtained through the semi-structured feature vector has the advantages of lower calculation amount, high running speed, low requirement on system hardware and low cost.

In an exemplary embodiment, the finding the vehicle passing through the camera includes: and searching for vehicles passing through the camera within a preset time period. The preset time period may be configured by a user or set by default by a system, and the preset time period may be a period of time before the vehicle is departed.

In an exemplary embodiment, the method further comprises prompting only the unlicensed vehicle to open the WiFi connection when the vehicle enters the field, i.e., the unlicensed vehicle opens the WiFi connection and the licensed vehicle closes the WiFi connection. According to the scheme provided by the embodiment, the WiFi connection with the licensed cars can be closed, the MAC information quantity in the WiFi MAC information total set can be reduced when the licensed cars exist in the second characteristic information sequence, and the WiFi MAC information set with the intersection can be conveniently searched.

In an exemplary embodiment, the method further comprises: and when the second characteristic information is matched with a plurality of entrance records and the second WiFi MAC information set only has intersection with the first WiFi MAC information set of one entrance record in the matched plurality of entrance records, charging the vehicles which come out according to the entrance time of the entrance records with the intersection. Generally, the number of the unlicensed vehicles in the parking lot is small, and the probability that the unlicensed vehicles with the same vehicle type enter the same parking lot is lower, so that only one admission record is matched, and the vehicles can be conveniently identified.

In an exemplary embodiment, the method further comprises: and when the target camera is not found, or the WiFi MAC information total set and a first WiFi MAC information set of more than one entry record in the matched entry records have intersection, prompting to carry out manual charging. The scheme provided by the embodiment can be used for charging the vehicles which cannot be confirmed in a manual charging mode. During manual charging, the historical video can be manually searched, and the entrance time is determined for charging.

In an exemplary embodiment, the prompting may be performed using one or a combination of the following: voice prompt and display screen prompt.

The technical solution of the embodiments of the present disclosure is further explained by an example.

According to the scheme provided by the embodiment, when a vehicle enters a field, the camera generates a corresponding semi-structured feature vector under a neural network model for a unlicensed vehicle, a WiFi Sniffer (WiFi Sniffer) module carried by the camera collects a WIFI MAC information set around the camera at the moment, and the feature vector and the WIFI MAC information set of the vehicle are sent to a background server; the background server takes the characteristic vector, the WiFi MAC information set and the entrance time as the entrance records of the vehicle and stores the entrance records into a database; if two similar unlicensed vehicles (vehicles with similar characteristic information) enter, the characteristic vectors and the WiFi MAC sets of different vehicles are stored separately.

When the vehicles leave, if no vehicle is detected, the cooperation background server judges whether other vehicles similar to the vehicle feature vector exist, if not (only one entry record is matched), charging processing is carried out according to the matched entry record, if yes (multiple entry records are matched), a WiFi MAC information set at the moment is collected, if the WiFi MAC information set only has intersection with the WiFi MAC information set matched with one of the multiple entry records, the vehicles can be identified, and charging is carried out according to the entry time of the vehicles.

If the collected WiFi MAC information set is intersected with the WiFi MAC information sets in the matched entrance records, which vehicle in the similar vehicles cannot be determined, calling at least one vehicle which goes out before, generating a feature vector sequence of the current vehicles which go out and the feature vectors of the called vehicles according to the order of the vehicles which go out, and identifying the vehicles according to the feature vector sequence.

FIG. 2 is a schematic diagram of a vehicle identification system provided in an exemplary implementation. As shown in fig. 2, the vehicle recognition system includes: the intelligent IPC monitoring system comprises an intelligent camera (or intelligent IPC), a display screen and a background server, wherein the background server is connected with the intelligent IPC (can be connected in a wired or wireless mode), the display screen is connected with the intelligent IPC (can be connected in a wired or wireless mode), or the display screen and the intelligent IPC can be integrated together, wherein:

the intelligent IPC can identify the license plate number and generate a semi-structured feature vector under a neural network model for the unlicensed vehicle. A WiFi MAC collector (namely a WiFi Sniffer module) can be integrated in the intelligent IPC, and wireless device information near the intelligent IPC, such as the MAC address of a mobile phone of a vehicle driver and a passenger, can be collected. The entrance and exit of the parking lot and each intersection inside can all install intelligent IPC.

The display screen may be placed at the entrance and exit of the parking lot. The display screen at the entrance can prompt the unlicensed vehicle to start WiFi (wireless fidelity) so as to authorize the WiFi Sniffer module to wirelessly collect WiFiMAC (wireless fidelity access control) of the mobile phone of the driver and the passenger, and can display the identified license plate number of the licensed vehicle; the display screen at the exit may prompt the parking time and the amount of the charge for the vehicle. In another exemplary embodiment, a display screen may not be provided, and voice prompt may be used.

The background server is configured to receive, store, retrieve, match the feature vectors uploaded by the intelligent IPC, and support analysis, filtering of WiFi MAC addresses provided by a WiFi Sniffer module of the intelligent IPC to identify the vehicle.

In another exemplary embodiment, the feature vector generation of the vehicle may be undertaken by a background server, and the intelligent IPC only needs to provide an image or video code stream of the vehicle, which is not limited by the scheme.

The license plate recognition method provided by the embodiment comprises the following steps:

step 301, after a vehicle enters a field, the intelligent IPC acquires image generation characteristic information and an MAC set;

when a vehicle enters a field, an intelligent camera at an entrance generates a semi-structured feature vector under a corresponding neural network model, namely a first feature vector, for the vehicle, WiFi Sniffer carried by the intelligent camera collects WiFi MAC information around the intelligent camera at the moment to generate a first WiFi MAC information set, and the first feature vector and the first WiFi MAC information set of the vehicle are sent to a database of a background server to be stored; and if two similar unlicensed vehicles enter, separately storing the vehicle feature vector, the WiFi MAC information set and the binding relationship.

The intelligent IPC at the entrance acquires the feature vector and WiFi MAC information set of the vehicle. After an interested target (such as a vehicle/a person) in a video image is subjected to target classification and target positioning, corresponding semi-structured feature vectors are extracted through a deep neural network. The feature vector can be a real number vector of 1 x 128 or 1 x 256, and different types of vehicle targets are mapped by the deep neural network to obtain different feature vectors. When a video image is collected, the intelligent IPC can also acquire surrounding WiFi MAC information through a WiFi Sniffer module. And the intelligent IPC reports the characteristic information, the Wifi MAC information set and the entrance time to the background server.

Supposing that a unlicensed vehicle A0 enters a park, the background server generates a record of an intelligent IPC snapshot vehicle according to information reported by the intelligent IPC at an entrance, and stores a snapshot time (namely, an entrance time), a feature vector and a Wifi MAC information set of the vehicle, which is called an entrance record.

As shown in table 1 below, the entry record may include: the vehicle identification information (namely the vehicle target ID) distributed to the vehicle by the background server, the IPC number (namely the IPC number reporting the characteristic information and the Wifi MAC information set), the unlicensed vehicle indication information (namely whether the unlicensed vehicle is indicated), the characteristic vector, the MAC information (namely the WiFi MAC information set) and the time of entry.

Table 1 entry record table one

Vehicle object ID	IPC numbering	License plate-free vehicle indication	Feature vector	MAC information	Time of admission
						RecordID1
	1	1	Feature vector A	a1，a2，c，d	ET1

In table 1, the actual MAC addresses of the cell phones of the occupants in the unlicensed car a0 are a1, a 2. And other MAC addresses c and d belong to mobile phones of other nearby people acquired by intelligent IPC at the entrance and are interference addresses.

The unlicensed vehicle indication field may be represented by 1 or 0, where a value of 1 indicates an unlicensed vehicle and a value of 0 indicates a licensed vehicle, which is merely an example here, and may be indicated in other manners, such as using letters, and so on.

In an exemplary embodiment, the unlicensed vehicle indication field and the vehicle target ID field may be removed.

Other intelligent IPCs in the field collect images of the vehicles to generate characteristic vectors and WiFi MAC when the vehicles pass by, report the characteristic vectors, the WiFi MAC when the vehicles pass by and image timestamps of the collected images of the vehicles (namely the timestamps of the images for generating the characteristic vectors) to the background server, and the background server generates records in the field. As shown in table 2, the intrafield records may include: the IPC number (namely the IPC number reporting the characteristic information and the Wifi MAC information set), the characteristic vector, the MAC information (namely the WiFi MAC information set) and the collection time.

TABLE 2 in-field record table

IPC numbering	License-plate-free vehicle	Feature vector	MAC information	Time of acquisition
					5	1	Feature vector A	a1，a2，e，f	ET5

Wherein, intelligent IPC in the field carries out the collection of eigenvector and MAC information to having the tablet vehicle and not having the tablet vehicle.

And step 302, when the unlicensed vehicle leaves, if the characteristic information of the unlicensed vehicle is only matched with one admission record, carrying out charging processing according to the admission time in the matched admission record.

When the unlicensed vehicle leaves the field, the intelligent IPC at the exit collects video images, and matches the collected characteristic vectors of the vehicle with the entrance records of the unlicensed vehicles which are not left in the background server (namely, the collected characteristic vectors are compared with the characteristic vectors in the entrance records, for example, the Euclidean distance or the cosine distance between the collected characteristic vectors and the entrance records is calculated). If only one entry record is matched through the feature vector matching, the parking fee is directly calculated according to the time difference between the departure time and the entry time of the matched entry record, and the parking fee is displayed on a display screen and is paid by a driver.

When the unlicensed vehicle leaves, if the characteristic information of the unlicensed vehicle is matched with a plurality of entrance records, acquiring the unlicensed vehicle at the current moment and a peripheral WiFi MAC information set (namely the peripheral WiFi MAC information set of the intelligent IPC at the place of the unlicensed vehicle), and when the WiFi MAC information set is intersected with only one of the matched entrance records, charging the current exit time according to the entrance time in the entrance records with the intersection.

When two similar vehicles enter the field, the intelligent IPC of the entrance will generate an entrance record for each of them, as shown in table 3 and table 4, respectively, and the feature vectors of both are the same or similar.

TABLE 3 entry record Table II

Vehicle object ID	IPC numbering	License-plate-free vehicle	Feature vector	MAC information	Time of admission
						RecordID1
	1	1	Feature vector A	a1，a2，c，d	ET1

Table 4 entry record table three

Vehicle object ID	IPC numbering	License-plate-free vehicle	Feature vector	MAC information	Time of admission
						RecordID9
	1	1	Feature vector A'	a9，e，f	ET9

Although the feature vector A and the feature vector A' are not completely the same, the Euclidean distance or the cosine distance between the two is less than or equal to the preset threshold value, so that the system can judge that the vehicles are the same type (similar).

When the unlicensed vehicle leaves, the intelligent IPC at the exit collects video images, determines the characteristic vector of the vehicle according to the collected images, and compares the characteristic vector with the characteristic vector of the unlicensed vehicle which is not left and stored in the background server. And matching 2 entry records through feature vector comparison, and acquiring WiFi MAC information sets around the moment by the intelligent IPC at the exit. And if the collected WiFi MAC information set only has intersection with the WiFi MAC information set of one of the matched 2 entry records, carrying out charging processing according to the entry time and the current exit time in the entry records with the intersection. For example, if the collected WiFi MAC information set is { a1, g, h }, and only an intersection exists with the entry record shown in table 3, it indicates that the vehicle is a vehicle corresponding to the entry record shown in table 3, and charging is performed according to the entry time and the current exit time in table 3.

And step 304, when the unlicensed vehicle leaves the field, if the characteristic information of the unlicensed vehicle is matched with a plurality of entrance records, acquiring the unlicensed vehicle at the current moment and a WiFi MAC information set around the intelligent IPC (International Mobile communication interface) (namely, a WiFi MAC information set around the intelligent IPC at the exit), and when the WiFi MAC information set intersects with the WiFi MAC information sets of more than one entrance record in the matched entrance records, identifying the vehicle according to the characteristic vector sequence of a plurality of continuous vehicles which leave the field.

That is, if the collected WiFi MAC information set intersects with the WiFi MAC information sets of the matched multiple entry records, it is impossible to determine which vehicle is a similar vehicle, call at least one vehicle that has gone out before, generate a feature vector sequence from feature vectors of the current vehicle and the called vehicle according to the order of departure, and send the sequence to multiple cameras at nearby intersections. If one camera acquires a vehicle with the same sequence, acquiring a WiFi MAC information total set acquired when the vehicle sequence of the camera passes by; and when the WiFi MAC information total set only has intersection with the WiFi MAC information set of one entry record in the matched entry records, charging according to the entry time in the entry record with the intersection.

A specific example is illustrated. If the exit camera acquires surrounding WiFi MAC information sets of { a1, a9, h, i }, and the eigenvector is a, there is an intersection with the two entry records, record id1 and record id9, which is the case because another similar vehicle is parked nearby. And the video camera at the exit returns the historical video, determines a plurality of vehicles, such as 2 vehicles, before the vehicle which is currently out of the field, such as C, B, A feature vectors of vehicles which are out of the field in time sequence respectively, and obtains a feature vector sequence C-B-A, wherein C is the feature vector of the vehicle which is the earliest out of the called vehicles, A is the feature vector of the vehicle which is currently out of the field, the vehicle corresponding to C-B-A is the vehicle which is continuously out of the field, namely the vehicle which is not out of the field between C and B, and the vehicle which is not out of the field between B and A. The camera at the exit sends the feature vector sequence to a plurality of cameras nearby, and the cameras retrieve the history videos within a preset time period (for example, 5 minutes, which can be set by a user or preset by a system, wherein the time period is set, firstly, the history videos are retrieved too far so that the peripheral cameras do not need to consume resources, and secondly, the probability that another similar vehicle participates in the similar sequence is increased when the time is too long). If one camera records through backtracking history, the vehicles with similar feature vector sequences pass through the own view according to the same sequence, for example, the vehicles with the feature vector sequences C '-B' -A '(or X1-C' -B '-X2-A'), X1 and X2 are vehicles which are the approach target cameras but are not present. Wherein C 'is similar to C, B' is similar to B, A 'is similar to A, and the vehicle corresponding to C', the vehicle corresponding to B 'and the vehicle corresponding to A' continuously pass through the camera, and the sequence of passage is consistent with the sequence of C, B, A exits, the camera is called the target camera, the target camera combines the WiFi MAC information set corresponding to C ', the WiFi MAC information set corresponding to B ' and the WiFi MAC information set corresponding to A ' into a WiFi MAC information total set, feeds the WiFi MAC information total set back to the camera at the outlet, for example, the WiFi MAC information aggregate is { c1, i, k } { b1, h, k } { a1, j, z }, this total set of WiFi MAC information intersects only the set of WiFi MAC information of recordids 1, and if the WiFi MAC information of the RecordID9 is not intersected, the current vehicle is the vehicle corresponding to the RecordID1, and charging according to the entry time and the current exit time of the entry record corresponding to the RecordID 1. In this embodiment, the number of vehicles that have been called and are going out of the field is 2, but the embodiment of the present disclosure is not limited thereto, and may be 1 or more than 2. The preset time period may be a period of time from the current time to the previous time.

In an exemplary embodiment, the background server may trace back the history videos, determine the feature vector sequence, trace back the history videos of other cameras, determine a similar feature vector sequence and a WiFi MAC information total set, and determine that there is an entrance record with an intersection, thereby implementing charging.

In an exemplary embodiment, a display screen at an entrance prompts that only the unlicensed vehicles start the WiFi, when the unlicensed vehicles possibly exist in the vehicle sequence, the licensed vehicles turn off the WiFi, and the corresponding MAC information fields do not exist, so that the WiFi MAC information sets of the vehicle sequence are combined in a smaller range, the WiFi MAC information sets with intersection are easily and quickly matched, the vehicles can be quickly identified, and the charging efficiency is improved.

As shown in fig. 3, the embodiment of the present disclosure provides a vehicle identification device 40, which includes a memory 410 and a processor 420, where the memory 410 stores a program, and when the program is read and executed by the processor 420, the program implements the vehicle identification method according to any one of the embodiments.

The disclosed embodiments provide a computer-readable storage medium storing one or more programs, which are executable by one or more processors to implement the vehicle identification method of any of the above embodiments.

The embodiment of the disclosure provides a vehicle identification system, including the camera in entrance, the camera in exit and at least one camera in the scene, and background server, can be connected through wired or wireless mode between background server and the camera, can be connected through wired or wireless mode between the camera, wherein:

the camera at the entrance is configured to acquire an image of the vehicle when an unlicensed vehicle enters the entrance, or acquire the image of the vehicle to generate first characteristic information, and send the image or the first characteristic information and the entrance time of the vehicle to the background server; acquiring wireless fidelity media access control information of the vehicle and the periphery, generating a first wireless fidelity media access control information set, and sending the first wireless fidelity media access control information set to the background server;

the camera at the exit is configured to acquire an image of a vehicle which is currently present when an unlicensed vehicle is present, or acquire second characteristic information generated by the vehicle which is currently present, send the image or the second characteristic information to the background server, acquire wireless fidelity media access control information of the vehicle which is currently present and the periphery, generate a second wireless fidelity media access control information set and send the second wireless fidelity media access control information set to the background server;

the cameras in the field are configured to acquire images of passing vehicles when the unlicensed vehicles pass by, or acquire the images of the vehicles to generate third feature information, and send the images or the third feature information and image timestamps of the images of the passing vehicles to the background server; acquiring wireless fidelity media access control information of the passing vehicles and the periphery to generate a third wireless fidelity media access control information set and sending the third wireless fidelity media access control information set to the background server;

the background server is configured to receive an image sent by a camera at an entrance to generate first characteristic information, or receive the first characteristic information sent by the camera at the entrance, receive a first wireless fidelity medium access control information set sent by the camera at the entrance, generate and store an entry record of a vehicle, wherein the entry record comprises the first characteristic information, the first wireless fidelity medium access control information set and the entry time of the vehicle;

receiving an image sent by a camera in a field to generate third characteristic information or receiving third characteristic information sent by the camera in the field, receiving a third wireless fidelity medium access control information set sent by the camera in the field, generating and storing a field record of the vehicle, wherein the field record comprises identification information of the camera in the field, the third characteristic information and the third wireless fidelity medium access control information set, and the camera acquires an image time stamp of the image of the passing vehicle;

receiving an image sent by the camera at the exit to generate second characteristic information or receiving the second characteristic information sent by the camera at the exit, and receiving a second wireless fidelity medium access control information set sent by the camera at the exit;

when the second characteristic information is matched with a plurality of entrance records and intersection exists between the second wireless fidelity media access control information set and a first wireless fidelity media access control information set of more than one entrance record in the plurality of matched entrance records, second characteristic information of at least one vehicle which is out of the field before the vehicle which is out of the field currently is obtained, and a second characteristic information sequence is generated according to the appearance sequence by the second characteristic information of the vehicle which is out of the field currently and the second characteristic information of the vehicle which is out of the field already. The vehicles corresponding to the second characteristic information in the second characteristic information sequence are a plurality of continuous vehicles which come out of the field in sequence;

searching vehicles passing through the cameras for a plurality of cameras preset in a field, and merging wireless fidelity media access control information sets corresponding to each piece of third feature information in a third feature information sequence according to a passing sequence to generate a wireless fidelity media access control information total set according to the field record of the target camera when one target camera exists, the target camera has the third feature information sequence formed by the third feature information of the passing vehicles according to the passing sequence, and the ith third feature information in the third feature information sequence is similar to the ith second feature information in the second feature information sequence; and when the total wireless fidelity media access control information set only has intersection with a first wireless fidelity media access control information set of one of the matched entry records, charging the vehicles on the departure according to the entry time of the entry record with the intersection, wherein the passing sequence is consistent with the time sequence of the departure sequence, i is 1 to n, n is the number of second feature information in the second feature information sequence, and n is greater than 1.

It will be understood by those of ordinary skill in the art that all or some of the steps of the methods, systems, functional modules/units in the devices disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. In a hardware implementation, the division between functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed by several physical components in cooperation. Some or all of the components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as is well known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media as is well known to those skilled in the art.

Claims (10)

1. A vehicle identification method, characterized by comprising:

when a unlicensed vehicle enters a field, acquiring an image of the vehicle to generate first characteristic information, acquiring wireless fidelity media access control information of the vehicle and the periphery to generate a first wireless fidelity media access control information set, generating and storing a field entry record of the vehicle, wherein the field entry record comprises the first characteristic information, the first wireless fidelity media access control information set and the field entry time of the vehicle; when a vehicle is in a field, generating third characteristic information according to an image of the vehicle acquired by a camera through which the vehicle passes, acquiring wireless fidelity media access control information of the vehicle and the periphery of the vehicle to generate a third wireless fidelity media access control information set, and generating and storing an in-field record of the vehicle, wherein the in-field record comprises identification information of the camera, the third characteristic information, the third wireless fidelity media access control information set and an image timestamp of the image of the vehicle acquired by the camera;

when the unlicensed vehicle leaves, acquiring an image of the vehicle which leaves at present to generate second characteristic information, and acquiring the vehicle which leaves at present and wireless fidelity media access control information of the periphery to generate a second wireless fidelity media access control information set; matching the second characteristic information with entry records, when the second characteristic information is matched with a plurality of entry records and the second wireless fidelity media access control information set is intersected with the first wireless fidelity media access control information set of more than one entry record in the matched entry records, acquiring the second characteristic information of at least one vehicle which is out of the field before the vehicle which is out of the field currently, generating a second characteristic information sequence by the vehicle which is out of the field currently and the second characteristic information of the vehicle which is out of the field currently according to the field-out sequence, wherein the vehicles corresponding to the second characteristic information in the second characteristic information sequence are a plurality of continuous vehicles which are out of the field sequentially;

searching vehicles passing through the cameras for a plurality of cameras preset in a field, and merging a wireless fidelity media access control information set corresponding to each piece of third feature information in a third feature information sequence according to the in-field record of a target camera to generate a wireless fidelity media access control information total set when the target camera exists and the target camera has the third feature information sequence formed by the third feature information of the passing vehicles according to the passing sequence and the ith third feature information in the third feature information sequence is similar to the ith second feature information in the second feature information sequence; and when the total wireless fidelity media access control information set only has intersection with a first wireless fidelity media access control information set of one of the matched entry records, charging the vehicles on departure according to the entry time of the entry record with the intersection, wherein the passing sequence is consistent with the time sequence of the departure sequence, i is 1 to n, n is the number of second characteristic information in the second characteristic information sequence, and n is greater than 1.

2. The vehicle recognition method according to claim 1, wherein the third feature information sequence is a sequence of third feature information of vehicles that continuously approach the target camera.

3. The vehicle identification method according to claim 1, wherein the searching for the vehicle passing through the camera includes: and searching for vehicles passing through the camera within a preset time period.

4. The vehicle recognition method according to claim 1,2 or 3, wherein the plurality of cameras preset in the field include a plurality of cameras near a departure place.

5. The vehicle identification method according to claim 1,2 or 3, characterized in that the method further comprises: and when the second characteristic information is matched with a plurality of entrance records and the second wireless fidelity media access control information set only has intersection with the first wireless fidelity media access control information set of one entrance record in the matched plurality of entrance records, charging the vehicles which come out of the vehicle according to the entrance time of the entrance record with the intersection.

6. The vehicle identification method according to claim 1,2 or 3, characterized in that the method further comprises: and when the second characteristic information is only matched with one entry record, charging the vehicles for leaving according to the entry time of the matched entry record.

7. The vehicle identification method according to claim 1,2 or 3, characterized in that the first feature information, the second feature information, and the third feature information are semi-structured feature vectors generated by a preset neural network model.

8. A vehicle identification device comprising a memory and a processor, the memory storing a program that, when read and executed by the processor, implements a vehicle identification method according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that the computer-readable storage medium stores one or more programs executable by one or more processors to implement the vehicle identification method according to any one of claims 1 to 7.

10. A vehicle identification system comprising a camera at an entrance, a camera at an exit and at least one camera within a field, and a background server, wherein:

the camera at the entrance is configured to acquire an image of a vehicle when the unlicensed vehicle enters the field, or acquire the image of the vehicle to generate first feature information, and send the image or the first feature information and the time of the vehicle entering the field to the background server; acquiring wireless fidelity media access control information of the vehicle and the periphery to generate a first wireless fidelity media access control information set and send the first wireless fidelity media access control information set to the background server;

the camera at the exit is configured to acquire an image of a vehicle which is currently present when an unlicensed vehicle is present, or acquire second feature information generated by the vehicle which is currently present, send the image or the second feature information to the background server, acquire wireless fidelity media access control information of the vehicle which is currently present and the periphery, generate a second wireless fidelity media access control information set and send the second wireless fidelity media access control information set to the background server;

the cameras in the field are configured to acquire images of passing vehicles when the unlicensed vehicles pass by, or acquire the images of the vehicles to generate third feature information, and send the images or the third feature information and image timestamps of the images acquired by the cameras to the background server; acquiring wireless fidelity media access control information of the passing vehicles and the periphery to generate a third wireless fidelity media access control information set and sending the third wireless fidelity media access control information set to the background server;

the background server is configured to receive an image sent by a camera at an entrance to generate first feature information, or receive the first feature information sent by the camera at the entrance and a first wireless fidelity media access control information set sent by the camera at the entrance, generate and store an entry record of a vehicle, wherein the entry record comprises the first feature information, the first wireless fidelity media access control information set and an entry time of the vehicle;

receiving an image sent by a camera in a field to generate third feature information or receiving third feature information sent by the camera in the field, receiving a third wireless fidelity media access control information set sent by the camera in the field, generating and storing a field record of the vehicle, wherein the field record comprises identification information of the camera in the field, the third feature information and the third wireless fidelity media access control information set, and the camera acquires an image time stamp of the image of the passing vehicle;

receiving an image sent by the camera at the exit to generate second characteristic information or receiving the second characteristic information sent by the camera at the exit, and receiving a second wireless fidelity medium access control information set sent by the camera at the exit;

when the second characteristic information is matched with a plurality of entrance records and the intersection exists between the second wireless fidelity media access control information set and a first wireless fidelity media access control information set of more than one entrance record in the plurality of matched entrance records, acquiring second characteristic information of at least one vehicle which is out of the field before the vehicle which is out of the field currently, and generating a second characteristic information sequence according to the appearance sequence by using the second characteristic information of the vehicle which is out of the field currently and the second characteristic information of the vehicle which is out of the field already; the vehicles corresponding to the second characteristic information in the second characteristic information sequence are a plurality of continuous vehicles which come out of the field in sequence;

searching vehicles passing through the cameras for a plurality of cameras preset in a field, and merging a wireless fidelity media access control information set corresponding to each piece of third feature information in a third feature information sequence according to the in-field record of a target camera to generate a wireless fidelity media access control information total set when the target camera exists and the target camera has the third feature information sequence formed by the third feature information of the passing vehicles according to the passing sequence and the ith third feature information in the third feature information sequence is similar to the ith second feature information in the second feature information sequence; and when the total wireless fidelity media access control information set only has intersection with a first wireless fidelity media access control information set of one of the matched entry records, charging the vehicles on departure according to the entry time of the entry record with the intersection, wherein the passing sequence is consistent with the time sequence of the departure sequence, i is 1 to n, n is the number of second characteristic information in the second characteristic information sequence, and n is greater than 1.

Images