CN111667697B

CN111667697B - Abnormal vehicle identification method and device, and computer readable storage medium

Info

Publication number: CN111667697B
Application number: CN201910167470.XA
Authority: CN
Inventors: 杨洪斌
Original assignee: Beijing Jingdong Century Trading Co Ltd; Beijing Jingdong Shangke Information Technology Co Ltd
Current assignee: Beijing Jingdong Century Trading Co Ltd; Beijing Jingdong Shangke Information Technology Co Ltd
Priority date: 2019-03-06
Filing date: 2019-03-06
Publication date: 2022-11-08
Anticipated expiration: 2039-03-06
Also published as: CN111667697A

Abstract

The disclosure relates to an abnormal vehicle identification method and device, and a computer readable storage medium. The abnormal vehicle identification method includes: acquiring a monitoring index of a vehicle to be detected; and judging whether the vehicle to be detected is an abnormal vehicle or not according to the vehicle monitoring index to be detected. According to the method and the system, under the premise that the case handling personnel do not receive any clues and do not know the case, the system can automatically identify the vehicle with abnormal behavior according to the monitoring data and early warn the vehicle possibly involved in the case, namely, the method and the system can realize no input but output.

Description

Abnormal vehicle identification method and device, and computer-readable storage medium

Technical Field

The present disclosure relates to the field of vehicle identification, and in particular, to a method and an apparatus for identifying an abnormal vehicle, and a computer readable storage medium.

Background

Convenient trip mode, unblocked highway brings the new challenge to city safety when bringing convenience for the city, and the criminal only needs several hours from the immigration to the province.

Disclosure of Invention

The applicant found that: in the related technology, when a case-related behavior and a virus-related behavior are detected, a suspected group is defined by adopting artificial analysis, then a technology-assisted examination mode is adopted, a monitoring system generates vehicle attribute information through video data of a monitoring camera, and a case handling person needs to speak a license plate number, a vehicle type and a vehicle color of a suspected vehicle to locate information and video data identified by the system. That is to say, each clue involved in a case still needs to be analyzed manually, the monitoring system can only screen results according to vehicle attributes, and cannot intelligently identify behavior before and after the case, so as to provide clues for case handling personnel, namely no input and no output are provided.

In the related technology, a mode of reporting a determined route by a person and then checking vehicles in a key road section one by one is adopted, and the time efficiency requirement of the required line reporting is too high and the good pursuit is finally missed because the vehicles required to be checked are too large and the police force required to be organized is too much.

The biggest defect of the related technology is that case-related clues need to be provided manually, and a monitoring system can only screen results according to attributes, only plays a role in auxiliary investigation and cannot automatically identify suspicious vehicles. Namely: there is no input and no output.

In view of the above technical problems, the present disclosure provides an abnormal vehicle identification method and apparatus, and a computer-readable storage medium, which can automatically identify an abnormally-behaving vehicle from monitored data without receiving any clue.

According to an aspect of the present disclosure, there is provided an abnormal vehicle identification method including:

acquiring a monitoring index of a vehicle to be detected;

and judging whether the vehicle to be detected is an abnormal vehicle or not according to the monitoring index of the vehicle to be detected.

In some embodiments of the present disclosure, the obtaining the monitoring index of the vehicle to be tested includes:

and determining the travel of the vehicle to be tested and the monitoring index of the vehicle to be tested according to the traffic information of the vehicle gate to be tested.

In some embodiments of the present disclosure, it is preferred, the passing information of the vehicle gate to be detected comprises at least one of gate codes, gate passing time and passing vehicle information of the gate.

In some embodiments of the present disclosure, the vehicle monitoring index to be measured includes at least one of a travel code, a travel time period, a transit time, a number of passing gates, vehicle travel data of vehicles passing through the same travel in a predetermined period, and vehicle travel data of vehicles passing through the same travel in the predetermined period and the same travel time period.

In some embodiments of the present disclosure, the abnormal vehicle is a vehicle involved in a case.

In some embodiments of the present disclosure, the determining whether the vehicle to be tested is an abnormal vehicle according to the vehicle monitoring index to be tested includes:

determining the travel of the vehicle to be tested;

determining a stroke route to be detected according to the stroke code of the vehicle to be detected;

and judging whether the vehicle to be detected is an abnormal vehicle or not according to the vehicle monitoring index to be detected aiming at the vehicle to be detected with the stroke code as the stroke route to be detected.

judging whether the vehicle to be detected has abnormal vehicle track conditions or not according to the monitoring indexes of the vehicle to be detected;

judging whether the abnormal traveling condition of the vehicle exists or not according to the monitoring index of the vehicle to be detected;

and determining whether the vehicle to be detected is an abnormal vehicle according to at least one of the abnormal vehicle track condition, the abnormal vehicle travel condition and the abnormal vehicle information clue of the vehicle to be detected.

In some embodiments of the present disclosure, the abnormal vehicle track condition includes at least one of an abnormal vehicle speed, an abnormal number of vehicles passing through a gate, and an abnormal vehicle transit time.

In some embodiments of the present disclosure, the abnormal vehicle travel condition includes at least one of an abnormal vehicle travel destination, an abnormal vehicle travel frequency, and an abnormal vehicle travel time.

In some embodiments of the present disclosure, the determining whether the vehicle to be tested has the abnormal vehicle track condition according to the monitoring index of the vehicle to be tested includes:

according to the passing time of vehicles passing through each gate in the same travel period, setting a plurality of passing standard time intervals passing through the gates for each gate;

comparing the passing time of the vehicle to be detected passing through each gate in the travel with a plurality of passing standard time intervals passing through the gate, and determining the passing cost of the vehicle to be detected passing through each gate, wherein the passing cost of the vehicle to be detected passing through each gate corresponds to the passing standard time interval in which the communication time of the vehicle to be detected passing through each gate is located;

determining the stability degree of the vehicle to be detected in the travel according to the mathematical variance;

and determining whether the vehicle to be detected in the travel has abnormal vehicle speed according to the stability degree of the passing cost of the vehicle to be detected in the travel.

and comparing the number of the bayonets passed by the vehicle to be detected in one stroke with the number of the bayonets passed by the vehicle in the same stroke and the same travel time period, and determining whether the number of the bayonets passed by the vehicle to be detected is abnormal.

and comparing the total passing time of the vehicle to be detected in one trip with the total passing time of the vehicle in the same trip and the trip time period, and determining whether the total passing time of the vehicle to be detected is abnormal.

In some embodiments of the present disclosure, the determining whether the vehicle to be tested has the abnormal vehicle track condition according to the monitoring index of the vehicle to be tested includes: and inputting the monitoring index of the vehicle to be detected into the vehicle track abnormity model to obtain the vehicle track abnormity condition.

In some embodiments of the present disclosure, the determining whether the abnormal vehicle travel condition exists in the vehicle to be detected according to the monitoring index of the vehicle to be detected includes: and inputting the monitoring indexes of the vehicle to be detected into the abnormal vehicle travel model to obtain the abnormal vehicle travel condition.

In some embodiments of the present disclosure, the determining whether the vehicle to be tested is an abnormal vehicle according to at least one of a vehicle track abnormal situation and a vehicle travel abnormal situation of the vehicle to be tested includes:

In some embodiments of the present disclosure, the determining whether the vehicle to be tested is an abnormal vehicle according to at least one of a vehicle track abnormal situation, a vehicle travel abnormal situation, and an abnormal vehicle information clue of the vehicle to be tested includes: and inputting at least one of the abnormal vehicle track condition, the abnormal vehicle travel condition and the abnormal vehicle information clue of the vehicle to be detected into the abnormal vehicle identification model, and determining whether the vehicle to be detected is an abnormal vehicle.

dividing the obtained vehicle monitoring indexes to be tested into training set data and test set data;

training an abnormal vehicle track recognition model by adopting training set data;

and inputting the test set data into the trained abnormal vehicle track recognition model, and determining whether the vehicle to be tested is an abnormal vehicle.

According to another aspect of the present disclosure, there is provided an abnormal vehicle recognition apparatus including:

the driving data acquisition module is used for acquiring a monitoring index of the vehicle to be detected;

and the abnormal vehicle identification module is used for judging whether the vehicle to be detected is an abnormal vehicle according to the vehicle monitoring index to be detected.

In some embodiments of the present disclosure, the abnormal vehicle recognition apparatus is configured to perform an operation of implementing the abnormal vehicle recognition method according to any one of the above embodiments.

a memory to store instructions;

a processor configured to execute the instructions to cause the abnormal vehicle identification apparatus to perform operations to implement the abnormal vehicle identification method according to any one of the above embodiments.

According to another aspect of the present disclosure, a computer-readable storage medium is provided, which stores computer instructions that, when executed by a processor, implement the abnormal vehicle identification method according to any one of the above embodiments.

The system can automatically identify the abnormal behavior vehicle according to the monitoring data and warn vehicles possibly involved in the case on the premise that the case handling personnel do not receive any clues and do not know the case, namely, the system can realize no input but output.

Drawings

In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without creative efforts.

FIG. 1 is a schematic diagram of some embodiments of an anomalous vehicle identification method of the present disclosure.

FIG. 2 is a schematic illustration of further embodiments of the disclosed abnormal vehicle identification method.

FIG. 3 is a schematic diagram of an abnormal vehicle trajectory identification model in some embodiments of the present disclosure.

FIG. 4 is a schematic diagram of a normal vehicle bayonet travel cost table in some embodiments of the present disclosure.

Fig. 5 is a schematic diagram of a card gate passing cost table of an abnormal vehicle (involved vehicle) in some embodiments of the disclosure.

FIG. 6 is a schematic diagram of a traffic cost table in some embodiments of the present disclosure.

FIG. 7 is a schematic diagram of a time cost table in some embodiments of the present disclosure.

FIG. 8 is a schematic illustration of further embodiments of the disclosed abnormal vehicle identification method.

FIG. 9 is a schematic illustration of a sample set in some embodiments of the disclosure.

FIG. 10 is a graphical representation of model training results in some embodiments of the present disclosure.

FIG. 11 is a schematic illustration of still further embodiments of the disclosed abnormal vehicle identification method.

FIG. 12 is a schematic diagram of a model index system in some embodiments of the present disclosure.

FIG. 13 is a graphical illustration of index weights in some embodiments of the disclosure.

FIG. 14 is a schematic illustration of still further embodiments of the disclosed abnormal vehicle identification method.

FIG. 15 is a schematic view of some embodiments of an anomalous vehicle identification device of the present disclosure.

FIG. 16 is a schematic view of further embodiments of the anomalous vehicle identification device of the present disclosure.

Detailed Description

The technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the drawings in the embodiments of the present disclosure, and it is obvious that the described embodiments are only a part of the embodiments of the present disclosure, and not all of the embodiments. The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the disclosure, its application, or uses. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present disclosure.

The relative arrangement of the components and steps, the numerical expressions, and numerical values set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.

Meanwhile, it should be understood that the sizes of the respective portions shown in the drawings are not drawn in an actual proportional relationship for the convenience of description.

Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate.

In all examples shown and discussed herein, any particular value should be construed as merely illustrative, and not limiting. Thus, other examples of the exemplary embodiments may have different values.

It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.

FIG. 1 is a schematic diagram of some embodiments of an anomalous vehicle identification method of the present disclosure. Preferably, the present embodiment may be executed by an abnormal vehicle recognition apparatus. The method comprises the following steps:

and 11, acquiring a monitoring index of the vehicle to be detected.

In some embodiments of the present disclosure, the obtaining of the monitoring index of the vehicle to be tested may include: and determining the travel of the vehicle to be tested and the monitoring index of the vehicle to be tested according to the traffic information of the vehicle gate to be tested.

The passing information of the vehicle gate to be detected comprises at least one of gate codes, gate passing time and passing vehicle information of the gate.

In some embodiments of the present disclosure, the traffic information of the vehicle to be tested may include camera data provided at a traffic light of a gate.

In some embodiments of the present disclosure, the determining the vehicle travel and the vehicle monitoring index according to the vehicle entrance traffic information to be measured includes: through carrying out preprocessing on the passing information of the gate, calculating a travel code and a monitoring index of the waiting-to-be-detected vehicle in a travel time period, and restoring a travel route of the vehicle.

In some embodiments of the present disclosure, the step of determining a monitoring index of the journey code waiting vehicle according to the gate traffic information may include:

step 11, it is assumed that the original bayonet traffic information data includes three adjacent lines of data as shown in table 1:

TABLE 1

Bayonet numbering	Passage time of gate	License plate for passing through bayonet
			001 intersection	22/2/2019	License plate number A0001
002 crossing	Year 2019, month 2, 23, day 5	License plate number A0001
			003 crossing	Year 2019, month 2, 23, day 10	License plate number A0001

And 12, calculating the time interval distribution between two gates according to the data of the vehicles passing through the gates, wherein an obvious time interval layer is required to be arranged between two adjacent intersections. From the 001 intersection to the 002 intersection, the vehicle takes a period of time divided into 3 hours and 10 hours (with a 7-hour break in between). Then, the vehicle that reaches the 002 intersection from the intersection 001 within 3 hours is described, and the 002 intersection is the gate through which the vehicle passes. The vehicle that passes through after 10 hours, which means that the 002 crossing is the first crossing (departure point) that it starts to pass through. The 001 intersection is the destination (end point) of the previous trip, that is, the vehicle returns to home after passing through the last gate in 2 months, 22 days, 30 days, the 001 intersection is the end point, and the 2 months, 23 days, the vehicle starts from home, reaches the first intersection 002 of the current trip at 5.

Step 13, a series of records connecting the starting point and the ending point are combined into a journey, and the journey comprises a departure point, an end point, a temporary trip, temporary trip time and the like, and subsequent journey costs, the number of journey passing temporary trips and the like are processed according to the table 1.

Step 14, adding a tolerance value to the duration of the discontinuous layer when the two intersections occur as a standard value of the slicing stroke, in this example, 3 hours +3 hours × 1.5=4.5 hours, if the vehicle from the intersection 001 to the intersection 002 exceeds 4.5 hours, 001 is the end point of the forward stroke, and 002 is the start point of the next stroke. If the time does not exceed 4.5 hours, the time is 002 the passing point of the stroke, and the stroke end needs to be continuously found.

In some embodiments of the present disclosure, the step of determining a monitoring index of the journey code waiting vehicle according to the gate traffic information may include: and determining the trip time period of the vehicle to be tested according to the passing information of the gate.

In some embodiments of the present disclosure, because the vehicle has peak value change of peak value in the morning and evening, whether the time spent between all two checkpoints is reasonable can be determined as abnormal, and the time needs to be considered in combination with the travel period.

And the travel time period is divided according to the vehicle peak value change in 24 hours in the city. Based on the following two principles:

first, the travel time period cannot be set too narrow, for example, a 24-hour day is divided into a time period 1-24, and in the time period 1, if only 3 vehicles pass through from the intersection a to the intersection B, respectively 20 minutes 30 minutes and 31 minutes, a sample is insufficient, and the 20-minute vehicle cannot be an abnormal vehicle with a large time difference from most of the vehicles in terms of variance,

Secondly, the trip time period cannot be set too long, and if the trip time period is set too long, the two peak time periods are wrapped, so that the identification accuracy of the abnormal vehicle is insufficient.

Therefore, in some embodiments of the present disclosure, the travel time period may be set to 3-5 time periods in 24 hours a day according to the morning and evening peaks and the morning and valley of the local city.

The method comprises the steps of identifying abnormal characteristics of each vehicle from a starting place to a shooting point at the moment of shooting by a camera, calculating abnormal values, determining where the vehicle comes from, where the vehicle goes to, what route the vehicle goes along, how much the vehicle drives, and identifying and marking suspicious points through sensitive places and other full-period characteristics.

And step 12, judging whether the vehicle to be detected is an abnormal vehicle or not according to the vehicle monitoring index to be detected.

In some embodiments of the present disclosure, the abnormal vehicle may be a vehicle involved in a case, such as a vehicle involved in a virus.

In some embodiments of the present disclosure, step 12 may comprise:

and step 121, determining a route (key route) of the to-be-detected vehicle according to the to-be-detected vehicle route code.

In some embodiments of the present disclosure, step 121 may comprise: firstly, the travel routes are grouped, and the route numbers with more cases (for example, more virus-related) are found to be used as the travel routes to be detected.

And step 122, judging whether the vehicle to be detected is an abnormal vehicle or not according to the vehicle to be detected monitoring index aiming at the vehicle to be detected with the stroke code as the stroke route to be detected.

Based on the abnormal vehicle identification method provided by the embodiment of the disclosure, on the premise that a case clerk does not receive any clues and does not know the occurrence of a case, the system can automatically identify the abnormal behavior vehicle according to the monitoring data and warn the vehicle possibly involved in the case, namely, the output without input can be realized.

FIG. 2 is a schematic illustration of further embodiments of the disclosed abnormal vehicle identification method. Preferably, the present embodiment may be executed by an abnormal vehicle recognition apparatus. The method (for example, the step of determining whether the vehicle to be tested is an abnormal vehicle according to the vehicle monitoring index in step 12 or step 122 in the embodiment of fig. 1) may include the following steps 21 to 23, where:

and step 21, judging whether the vehicle to be detected has the abnormal vehicle track condition according to the monitoring index of the vehicle to be detected.

In some embodiments of the present disclosure, step 21 may comprise: and inputting the monitoring index of the vehicle to be detected into the vehicle track abnormity model to obtain the vehicle track abnormity condition.

In some embodiments of the present disclosure, step 21 may comprise: correctly identifying a travel plan; and analyzing a series of checkpoints passed by the trip plan, including traffic speed analysis, communication starting point and destination analysis, trajectory topological structure analysis and the like.

FIG. 3 is a schematic diagram of an abnormal vehicle trajectory identification model in some embodiments of the present disclosure. The abnormal vehicle track identification model is used for judging whether the vehicle to be detected is an abnormal vehicle or not according to the vehicle monitoring index to be detected. As shown in fig. 3, the abnormal vehicle trajectory recognition model may include at least one of a vehicle trajectory abnormality model, a vehicle travel abnormality model, and an abnormal vehicle recognition model.

In some embodiments of the present disclosure, as shown in fig. 3, the abnormal vehicle track condition includes at least one of an abnormal vehicle speed (stop and pause), an abnormal vehicle passing through the gate (circle winding), and an abnormal vehicle passing time (abrasion and rub).

In some embodiments of the present disclosure, step 21 may identify an abnormal condition of abnormal vehicle speed (stop and pause), and the vehicle trajectory abnormality model in fig. 3 may be used to identify an abnormal condition of abnormal vehicle speed (stop and pause). At this time, step 21 may include:

and step 211, setting a plurality of passing standard time intervals passing through each gate for each gate according to the passing time of the passing vehicle passing through each gate in the same trip time period.

In some embodiments of the present disclosure, step 211 may include: and setting a bayonet passing cost table. Sequencing the passing time of the passing vehicles in the same time segment aiming at the vehicle route gate A and the next gate B thereof; the time spent on the 1/10 median is recorded as T1, the time spent on the 2/10 median is recorded as T2, and so on, and 9 reference values (T1, \8230;, T9) of the time spent on the crossing of the bayonet are obtained in turn.

And 212, comparing the passing time of the vehicle to be detected passing through each gate in the travel with a plurality of passing standard time intervals passing through the gate, and determining the passing cost of the vehicle to be detected passing through each gate, wherein the passing cost of the vehicle to be detected passing through each gate corresponds to the passing standard time interval of the communication time of the vehicle to be detected passing through each gate.

FIG. 4 is a schematic diagram of a normal vehicle bayonet travel cost table in some embodiments of the present disclosure. Fig. 5 is a schematic diagram of a card gate passing cost table of an abnormal vehicle (involved vehicle) in some embodiments of the disclosure.

In fig. 4, lines 1-7 are strokes 1 and lines 8-12 are strokes 2. Lines 1-11 in fig. 5 represent the 3 rd stroke. Each row in fig. 4 and 5 represents 9 reference values (T1, \8230;, T9) for a vehicle to pass through a gate, wherein the darker one is the transit time of the vehicle to be tested for the current trip through the gate. And determining the passing cost of the vehicle to be tested in the journey through the gate according to a reference value (T1, \8230; T9) corresponding to the passing time of the vehicle to be tested in the journey through the gate. For example: in fig. 4, lines 1-6 are the passing costs of the vehicle to be tested passing through the 1 st to 6 th bayonets in the 1 st journey, which are respectively (1, 2, 1).

And step 213, determining the stability of the running cost of the vehicle to be detected in the travel according to the mathematical variance.

And 214, determining whether the vehicle to be detected in the travel has abnormal vehicle speed according to the stability degree of the passing cost of the vehicle to be detected in the travel.

For the vehicle to be tested in the embodiment of fig. 4, the passing time of each gate is on the left side of the median, and the vehicle speed is stable; the vehicle is in a driving state without any contraindication, and therefore the vehicle to be tested in fig. 4 is a normal vehicle.

For the vehicle under test of the embodiment of fig. 5, on the same road segment, it is faster than most people to drive; the performance is very unstable after a while and slower than that of most people; is a driving state with concern; the system can be opened quickly, but the external environment needs to be observed by taking care. Therefore, the vehicle to be tested in fig. 5 is an abnormal vehicle (e.g., a vehicle involved in a case).

In some embodiments of the disclosure, for each journey, the passing cost of each gate in the journey is calculated, and the stability degree of the passing cost is represented by a mathematical variance. If the speed is stable and is always in a state of being faster or slower than others, the variance is minimum; this trip may be problematic if the speed is sometimes faster than others and sometimes slower than others.

In some embodiments of the present disclosure, step 21 may identify an abnormal condition that the vehicle passes through the abnormal number of gates (circle winding), and the vehicle track abnormal model in fig. 3 may be used to identify an abnormal condition that the vehicle passes through the abnormal number of gates (circle winding). At this time, step 21 may include: and comparing the number of the bayonets passed by the vehicle to be detected in one stroke with the number of the bayonets passed by the vehicle in the same stroke and travel time period, and determining whether the number of the bayonets passed by the vehicle to be detected is abnormal.

In some embodiments of the present disclosure, step 21 may comprise: determining vehicles in the same trip and same trip time period, wherein the first traffic light passing in the trip is the same, the last traffic light is the same, and the time period passing through the first traffic light is the same to be used as the vehicles in the same trip and same trip time period; sequencing the quantity of the bayonets passed by the vehicles in the same travel and same trip time period from large to small, and dividing the bayonets into 9 sections (9 grades), wherein the bayonets pass through the 1 grade at least, and the bayonets pass through the 9 grades at the most, and the grade is taken as the passing cost; and defining the vehicles represented by the deviated group as abnormal vehicles according to the discrete degree of the vehicle group to be detected and the vehicles in the same trip time period.

FIG. 6 is a schematic diagram of a traffic cost table in some embodiments of the present disclosure. Each row as shown in fig. 6 identifies a trip; presetting the number of bayonets of 9 gears for each stroke in a traffic cost table; and the darker one is the number of the bayonets passed by the vehicle to be tested in the journey.

As is apparent from the data in fig. 6, the trip 2 is a vehicle trip path when a case (e.g., a poison) is caught, and it can be seen that: when the drug-related personnel do not carry drugs, the shortest path is selected, and the traffic cost is the minimum. Once carrying drugs, the drug-related personnel need to take comprehensive consideration, and often detour in the same departure place and destination, compared with the vehicles (with the same departure point and destination), the number of passing bayonets is obviously more than the normal value, and the traffic cost index reflects the phenomenon of detour by comparing the number of passing bayonets with the traffic cost comparison table.

In the above embodiments of the present disclosure, the involved personnel tend to select the shortest path when not doing a case, and the traffic cost is the minimum. In case of crime, the involved personnel need to comprehensively consider, and in the same departure place and destination, the involved personnel need to bypass the inspection opening for avoiding inspection, so that compared with vehicles in the same journey (the same departure point, the same purpose and the same trip time period), the number of passing bayonets is obviously more than the normal value.

In some embodiments of this disclosure, step 21 can discern the abnormal conditions of vehicle transit time anomaly (rub-and-rub), and the abnormal conditions of vehicle orbit model in fig. 3 can be used for discerning the vehicle transit time anomaly (rub-and-rub). At this time, step 21 may include: and comparing the total passing time of the vehicle to be detected in one trip with the total passing time of the same trip and the same trip time period, and determining whether the total passing time of the vehicle to be detected is abnormal.

In some embodiments of the present disclosure, step 21 may comprise: determining vehicles in the same trip and same trip time period, wherein the first traffic light passing in the trip is the same, the last traffic light is the same, and the time period passing through the first traffic light is the same to be used as the vehicles in the same trip and same trip time period; sorting the time spent by vehicles in the same trip and same trip time period from large to small, and dividing the time into 9 sections (9 grades), wherein the least spent time is 1 grade, and the most spent time is 9 grades, and the grading is used as the time cost; and defining the vehicle with the largest time cost as an abnormal vehicle.

FIG. 7 is a schematic diagram of a time cost table in some embodiments of the present disclosure. Line 3 (bar 3) in fig. 7 is the vehicle trajectory when grasped.

In the above embodiment, due to detour and psychological pressure, the vehicle involved in a case may pass through more gates and spend more time in the total running process and wait for more favorable passing time, all of which are represented as abrasion and rubbing, compared with the vehicle in the same journey (same starting point-destination-time period) during the process of a crime.

And step 22, judging whether the vehicle to be detected has abnormal vehicle traveling conditions according to the vehicle monitoring index to be detected.

In some embodiments of the present disclosure, step 22 may comprise: and inputting the monitoring indexes of the vehicle to be detected into the abnormal vehicle travel model to obtain the abnormal vehicle travel condition.

In some embodiments of the present disclosure, as shown in fig. 3, the abnormal vehicle traveling condition may include at least one of an abnormal vehicle traveling destination condition, an abnormal vehicle traveling frequency condition, and an abnormal vehicle traveling time condition.

In some embodiments of the present disclosure, as shown in fig. 3, the abnormal vehicle travel condition may include at least one of an abnormal condition of dense treading, dense trading, dense joints, no wind noise leakage, and the like.

In some embodiments of the present disclosure, step 22 may comprise:

step 221, identifying the intensive treading point condition, and determining whether the intensive treading point condition occurs for multiple times in different time periods at the same destination within a predetermined statistical period, for example: stepping on the spot for multiple times respectively at morning peak, evening peak, morning, noon and evening.

Step 222, identifying the dense transaction situation, and determining whether the destination of the trip appears in the sensitive fence for many times within a predetermined statistical period.

And step 223, identifying the dense joint condition, and judging whether the destination appears in hotels, bars, entertainment places such as the hotel, the bar and the like, and places such as complex crowd places where public security events easily occur for many times within a preset counting period.

In some embodiments of the present disclosure, the predetermined statistical period may be 2-7 days.

And step 23, determining whether the vehicle to be detected is an abnormal vehicle according to at least one of the abnormal vehicle track condition, the abnormal vehicle travel condition and the abnormal vehicle information clue of the vehicle to be detected.

In some embodiments of the present disclosure, step 23 may comprise: and determining whether the vehicle to be detected is an abnormal vehicle according to at least one of the abnormal vehicle track condition, the abnormal vehicle travel condition and the abnormal vehicle information clue of the vehicle to be detected.

In some embodiments of the present disclosure, step 23 may comprise: and inputting at least one of the abnormal vehicle track condition, the abnormal vehicle travel condition and the abnormal vehicle information clue of the vehicle to be detected into the abnormal vehicle identification model, and determining whether the vehicle to be detected is an abnormal vehicle.

In some embodiments of the present disclosure, the abnormal vehicle information clues may include the value of the vehicle, the age, the model of the vehicle, driver information, and other clues derived from information.

In some embodiments of the present disclosure, step 23 may include identifying the involved vehicles based on the vehicle track abnormality model and the vehicle travel abnormality model, taking the involved vehicles in the past at the checkpoint as seed cases, and combining with the case transaction information clues.

In some embodiments of the present disclosure, the involved vehicles may include border-involved vehicles, provincial-involved vehicles, trans-provincial-involved vehicles, and the like.

In some embodiments of the present disclosure, the involved vehicles may include border-crossing, provincial, trans-provincial, freight, and passenger involved vehicles, among others.

According to the embodiment of the disclosure, the monitored vehicle information clues are brought into the abnormal condition of the vehicle track and the abnormal condition of the vehicle travel, so that the result is output.

The embodiment of the disclosure can automatically identify suspicious vehicles according to the physical attributes of the vehicle, such as speed, time consumption, driving track and the like, so that under the condition of no human input, from the viewpoint of automatically identifying the vehicles involved in the case by a machine, more optimized and upgraded schemes are provided in the future.

FIG. 8 is a schematic illustration of further embodiments of the disclosed abnormal vehicle identification method. Preferably, the present embodiment may be executed by an abnormal vehicle recognition apparatus. The method (for example, the step of determining whether the vehicle to be tested is an abnormal vehicle according to the vehicle monitoring index in step 12 or step 123 in the embodiment of fig. 1) may include the following steps:

and step 81, dividing the obtained vehicle monitoring indexes to be tested into training set data and test set data.

Step 82, training an abnormal vehicle track recognition model by using the training set data, wherein as shown in fig. 3, the abnormal vehicle track recognition model may include at least one of a vehicle track abnormal model, a vehicle travel abnormal model, and an abnormal vehicle recognition model.

And 83, inputting the test set data into the trained abnormal vehicle track recognition model, and determining whether the vehicle to be tested is an abnormal vehicle.

FIG. 9 is a schematic illustration of a sample set in some embodiments of the disclosure. As shown in fig. 9, the data of 7 and 8 months in a certain year are taken for testing: 15 hundred million pieces of bayonet data in 7 months and 8 months are merged into 71799 strokes. The data (sample set) was divided into training set 50214 and test set 21585, containing 157 and 69 actual involved trips, respectively. The percentage of the involved vehicles is three per thousand.

As shown in fig. 9, the vehicle monitoring indexes to be detected may include a license plate number, a transit code, a time period of a starting gate, a total transit duration, a total number of passing gates, a time variance spent by each gate, a transit cost, a number of vehicles on the same trip, a number of vehicles on the same track and at the same time point, a time cost, a number of times that the transit cost in the trip exceeds 90% of the median, and a number of times that the transit cost in the trip exceeds 10% of the median.

FIG. 10 is a graphical representation of model training results in some embodiments of the present disclosure. As shown in fig. 10, after training, verification is performed in the test set, and the test samples comprise 21654 strokes, 69 cases involved in the case, and the cases involved in the case accounts for three thousandths.

Outputting an abnormal vehicle track identification model: 22 items are output, 15 items are real cases, the accuracy rate is 68%, compared with the natural probability, the accuracy rate is improved by 219 times, and the model effect is obvious.

The biggest characteristic of the embodiment of the disclosure is that the machine automatically learns and excavates possible involved vehicles according to the psychological characteristics of the criminal and necessary precautionary measures without manually providing the involved vehicle clues as input conditions.

FIG. 11 is a schematic illustration of still further embodiments of the disclosed abnormal vehicle identification method. Preferably, the present embodiment may be executed by an abnormal vehicle recognition apparatus. As shown in fig. 11, the abnormal vehicle identification method may include:

and step 31, marking a sample.

In some embodiments of the present disclosure, step 31 may comprise: and processing historical vehicle checkpoint data.

In some embodiments of the present disclosure, step 31 may comprise: and (3) carrying out data cleaning, data insights and index labeling on the historical vehicle checkpoint data, judging whether the vehicle is involved in a case or not, determining the proportion of positive and negative samples, determining a sample set and the like.

And step 32, training the model.

In some embodiments of the present disclosure, step 32 may comprise: optimizing the variable number of node branches and the number of decision trees in a random forest, performing data missing processing, index type modification, data normalization, spatial interpolation, feature selection, index weight, evaluation and inspection and the like.

In some embodiments of the present disclosure, step 32 may comprise: the travel routes are grouped to find the serial numbers of the routes with more virus, and then the important routes are repeated to use indexes to carry out supervised and unsupervised learning.

In some embodiments of the present disclosure, the abnormal vehicle trajectory identification model may include an abnormal travel feature model.

FIG. 12 is a schematic representation of a model index system in some embodiments of the present disclosure. As shown in fig. 12, the model index system of the abnormal travel feature model may include: the system comprises the travel number, travel time periods, passing time, the number of passing gates, time variance spent by each gate, passing cost, time cost, vehicles passing through the same travel in a preset period, vehicles passing through the same travel in the same time period in the preset period and other indexes.

In some embodiments of the present disclosure, the predetermined period may be 2 months.

FIG. 13 is a graphical illustration of index weights in some embodiments of the disclosure. As shown in fig. 13, different model indices in fig. 12 correspond to different weights.

And step 33, predicting data.

In some embodiments of the present disclosure, step 33 may comprise: and processing the data of the newly added vehicle gate.

In some embodiments of the present disclosure, step 33 may comprise: and (4) carrying out data cleaning, data missing processing, data normalization and space interpolation on the newly added vehicle gate data, predicting the probability of case involvement, and screening the list according to the score.

The method of the embodiment of the disclosure can identify the abnormal characteristics of each vehicle from the departure place to the shooting point at the moment of shooting by the camera, calculate the abnormal value, determine where the vehicle comes from and goes to and where the vehicle goes, what route the vehicle travels, how much the vehicle speed is driven, and what sensitive and equal full-cycle characteristics pass through to identify and mark suspicious points.

FIG. 14 is a schematic illustration of further embodiments of the disclosed abnormal vehicle identification method. Preferably, the present embodiment may be executed by an abnormal vehicle recognition apparatus. As shown in fig. 14, the abnormal vehicle recognition method may include:

step 401, data preparation is performed.

Step 402, data processing is performed, for example: and generating a time cost table, a passing cost table and a passing cost table.

Step 403, exploratory data analysis is performed.

And step 404, analyzing historical vehicle gate indexes.

In step 405, a sample vehicle is selected.

And 406, performing sample representative analysis and rectification.

Step 407, performing travel behavior analysis.

Step 408, feature selection is performed.

Step 409, model training is performed.

In some embodiments of the present disclosure, step 409 may comprise step 82 of the fig. 8 embodiment or step 32 of the fig. 11 embodiment.

And step 410, performing data prediction and list output.

In some embodiments of the present disclosure, step 410 may include steps 21-23 of the fig. 2 embodiment, step 12 of the fig. 1 embodiment, step 83 of the fig. 8 embodiment, or step 33 of the fig. 11 embodiment.

Step 411, a capture scheme is specified.

At step 412, a capture scenario is executed.

And step 413, feeding back the capture effect.

And step 414, analyzing the optimized nodes so as to update the capture scheme in the step 411, perform optimization updating on the requirements in the step 403, and perform iterative optimization on the model.

FIG. 15 is a schematic view of some embodiments of an anomalous vehicle identification device of the present disclosure. As shown in fig. 15, the abnormal vehicle recognition apparatus of the present disclosure may include a traveling data acquisition module 151 and an abnormal vehicle recognition module 152, wherein:

and the driving data acquisition module 151 is used for acquiring a monitoring index of the vehicle to be detected.

In some embodiments of the present disclosure, the vehicle monitoring index to be measured may include a travel code, a travel time period, a transit time, a number of passing gates, vehicle travel data of vehicles passing through the same travel in a predetermined period, and vehicle travel data of vehicles passing through the same travel and the same travel time period in the predetermined period.

And the abnormal vehicle identification module 152 is configured to determine whether the vehicle to be detected is an abnormal vehicle according to the vehicle to be detected monitoring index.

In some embodiments of the present disclosure, the abnormal vehicle may be an involved vehicle.

In some embodiments of the present disclosure, the abnormal vehicle identification module 152 may be used to determine the trip of the vehicle under test; determining a route of the to-be-detected vehicle according to the to-be-detected vehicle route code; and judging whether the vehicle to be detected is an abnormal vehicle or not according to the vehicle monitoring index to be detected aiming at the vehicle to be detected with the stroke code as the stroke route to be detected.

In some embodiments of the present disclosure, the abnormal vehicle identification module 152 may be configured to determine whether the vehicle to be detected has a vehicle track abnormal condition according to the vehicle to be detected monitoring index; judging whether the abnormal traveling condition of the vehicle exists or not according to the monitoring index of the vehicle to be detected; and determining whether the vehicle to be detected is an abnormal vehicle according to at least one of the abnormal vehicle track condition, the abnormal vehicle travel condition and the abnormal vehicle information clue of the vehicle to be detected.

In some embodiments of the present disclosure, the vehicle track abnormality condition includes at least one of an abnormality in vehicle speed, an abnormality in the number of vehicles passing through a gate, and an abnormality in vehicle transit time.

In some embodiments of the present disclosure, the abnormal vehicle travel condition includes at least one of an abnormality of a vehicle travel destination, an abnormality of a vehicle travel frequency, and an abnormality of a vehicle travel time.

In some embodiments of the present disclosure, the abnormal vehicle identification module 152 may be configured to set, for each gate, a plurality of passage standard time intervals passing through the gate according to a passage time of a passing vehicle passing through each gate in the same passage time period; comparing the passing time of the vehicle to be detected passing through each gate in the travel with a plurality of passing standard time intervals passing through the gate, and determining the passing cost of the vehicle to be detected passing through each gate, wherein the passing cost of the vehicle to be detected passing through each gate corresponds to the passing standard time interval in which the communication time of the vehicle to be detected passing through each gate is located; determining the stability degree of the vehicle to be detected in the travel according to the mathematical variance; and determining whether the vehicle to be detected in the travel has abnormal vehicle speed according to the stability degree of the passing cost of the vehicle to be detected in the travel.

In some embodiments of the present disclosure, the abnormal vehicle identification module 152 may be configured to compare the number of passing bayonets of the vehicle to be tested in one trip with the number of passing bayonets of the vehicle in the same trip and the same trip time period, and determine whether there is an abnormality in the number of passing bayonets of the vehicle to be tested.

In some embodiments of the present disclosure, the abnormal vehicle identification module 152 may be configured to compare the total passing time of the vehicle to be tested in one trip with the total passing time of the same trip and the same trip time period, and determine whether there is an abnormality in the total passing time of the vehicle to be tested.

In some embodiments of the disclosure, the abnormal vehicle identification module 152 may be configured to input the vehicle monitoring index to be tested into the vehicle track abnormality model to obtain the vehicle track abnormality condition.

In some embodiments of the present disclosure, the abnormal vehicle identification module 152 may be configured to input the vehicle monitoring index to be tested into the vehicle travel abnormality model to obtain the vehicle travel abnormality.

In some embodiments of the present disclosure, the abnormal vehicle identification module 152 may be configured to determine whether the vehicle to be tested is an abnormal vehicle according to at least one of a vehicle track abnormal situation, a vehicle travel abnormal situation, and an abnormal vehicle information clue of the vehicle to be tested.

In some embodiments of the present disclosure, the abnormal vehicle identification module 152 may be configured to input at least one of a vehicle track abnormal situation, a vehicle travel abnormal situation, and an abnormal vehicle information clue of the vehicle to be tested into the abnormal vehicle identification model, and determine whether the vehicle to be tested is an abnormal vehicle.

In some embodiments of the present disclosure, the abnormal vehicle identification module 152 may be configured to divide the acquired vehicle monitoring indicators to be tested into training set data and test set data; training an abnormal vehicle track recognition model by adopting training set data; and inputting the test set data into the trained abnormal vehicle track recognition model, and determining whether the vehicle to be tested is an abnormal vehicle.

In some embodiments of the disclosure, the abnormal vehicle identification apparatus is configured to perform operations for implementing the abnormal vehicle identification method according to any one of the embodiments (for example, any one of fig. 1 to 14).

FIG. 16 is a schematic view of an abnormal vehicle identification apparatus of the present disclosure in accordance with still other embodiments. As shown in fig. 16, the disclosed abnormal vehicle recognition apparatus may include a memory 161 and a processor 162, wherein:

a memory 161 for storing instructions.

A processor 162 configured to execute the instructions, so that the abnormal vehicle identification apparatus performs the operation of implementing the abnormal vehicle identification method according to any one of the embodiments (for example, any one of fig. 1 to 14).

The method of the embodiment of the disclosure can identify the abnormal features of each vehicle from the departure place to the shooting point at the moment of shooting by the camera, calculate the abnormal value, determine where the vehicle comes from, where the vehicle goes to, what route the vehicle goes along, how much the vehicle speed is driven, and what sensitive places the vehicle passes through, and the like, and identify and mark suspicious points.

According to another aspect of the present disclosure, a computer-readable storage medium is provided, which stores computer instructions, which when executed by a processor, implement the abnormal vehicle identification method according to any one of the embodiments (for example, any one of fig. 1 to 14) described above.

The biggest characteristic of the embodiment of the disclosure is that the possible involved vehicles are automatically learned and excavated by a machine according to the psychological characteristics of the criminal and necessary precautionary measures without manually providing the involved vehicle clues as input conditions.

The abnormal vehicle identification apparatus described above may be implemented as a general purpose processor, a Programmable Logic Controller (PLC), a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any suitable combination thereof, for performing the functions described herein.

Thus far, the present disclosure has been described in detail. Some details that are well known in the art have not been described in order to avoid obscuring the concepts of the present disclosure. It will be fully apparent to those skilled in the art from the foregoing description how to practice the presently disclosed embodiments.

It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware to implement the above embodiments, where the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk, an optical disk, or the like.

The description of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the disclosure in the form disclosed. Many modifications and variations will be apparent to practitioners skilled in this art. The embodiment was chosen and described in order to best explain the principles of the disclosure and the practical application, and to enable others of ordinary skill in the art to understand the disclosure for various embodiments with various modifications as are suited to the particular use contemplated.

Claims

1. An abnormal vehicle identification method, characterized by comprising:

under the condition that no abnormal vehicle clue exists, acquiring a vehicle monitoring index to be detected;

judging whether the vehicle to be detected is an abnormal vehicle or not according to the vehicle monitoring index to be detected;

wherein, judge according to the vehicle monitoring index that awaits measuring whether the vehicle that awaits measuring is unusual vehicle includes:

judging whether the vehicle to be detected has abnormal vehicle track conditions according to the vehicle monitoring indexes to be detected;

wherein, whether the vehicle that awaits measuring has the vehicle orbit abnormal conditions according to the vehicle monitoring index that awaits measuring includes:

inputting the monitoring index of the vehicle to be detected into a vehicle track abnormity model to obtain the vehicle track abnormity condition, wherein the vehicle track abnormity model is used for identifying the abnormity condition of vehicle speed abnormity; at this time, the judging whether the vehicle to be detected has the abnormal condition of the vehicle track according to the monitoring index of the vehicle to be detected comprises the following steps:

according to the passing time of the passing vehicles passing through each gate in the same trip time period, setting a plurality of passing standard time intervals passing through the gate for each gate, wherein the plurality of passing standard time intervals are obtained by sequencing the passing time of the passing vehicles in the same time segment;

comparing the passing time of the vehicle to be detected passing through each gate in the travel with a plurality of passing standard time intervals passing through the gate, and determining the passing cost of the vehicle to be detected passing through each gate, wherein the passing cost of the vehicle to be detected passing through each gate is a reference value corresponding to the passing standard time interval in which the passing time of the vehicle to be detected passing through each gate is located;

determining the stability degree of the passing cost of the vehicle to be detected in the travel according to the mathematical variance of the passing cost of the vehicle to be detected passing through each gate;

and determining whether the vehicle to be detected in the travel has abnormal vehicle speed according to the stability of the passing cost of the vehicle to be detected in the travel.

2. The abnormal vehicle identification method according to claim 1, wherein the obtaining of the monitoring index of the vehicle to be tested comprises:

3. The abnormal vehicle identifying method according to claim 2,

the passing information of the vehicle gate to be detected comprises at least one of gate codes, gate passing time and gate passing vehicle information;

the vehicle monitoring indexes to be detected comprise at least one of travel codes, travel time periods, passing time, passing gate numbers, vehicle running data of vehicles passing through the same travel in a preset period, and vehicle running data of vehicles passing through the same travel and the same travel time periods in the preset period;

the abnormal vehicle is a vehicle involved in a case.

4. The abnormal vehicle identification method according to claim 3, wherein the determining whether the vehicle to be tested is an abnormal vehicle according to the vehicle monitoring index further comprises:

5. The abnormal vehicle identification method according to any one of claims 1 to 4, wherein after determining whether the vehicle to be tested has the vehicle track abnormal condition according to the vehicle monitoring index to be tested, the method further comprises:

judging whether the vehicle to be detected has abnormal vehicle traveling conditions according to the vehicle monitoring index to be detected;

determining whether the vehicle to be detected is an abnormal vehicle or not according to the vehicle track abnormal condition and the vehicle travel abnormal condition of the vehicle to be detected, wherein the vehicle track abnormal condition comprises abnormal vehicle speed.

6. The abnormal vehicle identifying method according to claim 5,

the abnormal condition of the vehicle track further comprises at least one of abnormal quantity of vehicles passing through the gate and abnormal vehicle passing time;

the abnormal vehicle traveling condition comprises at least one of abnormal vehicle traveling destination, abnormal vehicle traveling frequency and abnormal vehicle traveling time.

7. The abnormal vehicle identification method according to claim 6, wherein the step of judging whether the vehicle track abnormal condition exists in the vehicle to be detected according to the vehicle monitoring index further comprises the following steps:

8. The abnormal vehicle identification method according to claim 5, wherein the determining whether the vehicle to be detected is an abnormal vehicle according to the abnormal vehicle track condition and the abnormal vehicle travel condition of the vehicle to be detected comprises:

and determining whether the vehicle to be detected is an abnormal vehicle according to the abnormal vehicle track condition of the vehicle to be detected, the abnormal vehicle travel condition and the abnormal vehicle information clue.

9. The abnormal vehicle identifying method according to claim 8,

the step of judging whether the abnormal vehicle traveling condition exists in the vehicle to be detected according to the vehicle monitoring index to be detected comprises the following steps: inputting the monitoring indexes of the vehicle to be detected into a vehicle travel abnormity model to obtain vehicle travel abnormity conditions;

the step of determining whether the vehicle to be detected is an abnormal vehicle according to the abnormal vehicle track condition, the abnormal vehicle travel condition and the abnormal vehicle information clue of the vehicle to be detected comprises the following steps: and inputting the vehicle track abnormal condition, the vehicle travel abnormal condition and the abnormal vehicle information clue of the vehicle to be detected into the abnormal vehicle identification model, and determining whether the vehicle to be detected is an abnormal vehicle.

10. The abnormal vehicle identification method according to any one of claims 1 to 4, wherein the determining whether the vehicle under test is an abnormal vehicle according to the vehicle under test monitoring index includes:

training an abnormal vehicle track recognition model by adopting training set data, wherein the abnormal vehicle track recognition model comprises a vehicle track abnormal model;

11. An abnormal vehicle recognition device characterized by comprising:

the driving data acquisition module is used for acquiring a monitoring index of the vehicle to be detected under the condition that no abnormal vehicle clue exists;

the abnormal vehicle identification module is used for judging whether the vehicle to be detected is an abnormal vehicle or not according to the vehicle monitoring index to be detected;

the abnormal vehicle identification module is used for judging whether the vehicle to be detected has a vehicle track abnormal condition or not according to the vehicle monitoring index to be detected;

the abnormal vehicle identification module is used for inputting a vehicle monitoring index to be detected into a vehicle track abnormal model to obtain a vehicle track abnormal condition, wherein the vehicle track abnormal model is used for identifying an abnormal condition of abnormal vehicle speed; at the moment, the abnormal vehicle identification module is used for setting a plurality of passing standard time intervals passing through each gate aiming at each gate according to the passing time of the passing vehicles passing through each gate in the same passing time period under the condition of judging whether the vehicle track abnormal condition exists in the vehicles to be detected according to the vehicle monitoring indexes, wherein the plurality of passing standard time intervals are obtained by sequencing the passing time of the passing vehicles in the same time slice; comparing the passing time of the vehicle to be detected passing through each gate in the travel with a plurality of passing standard time intervals passing through the gate, and determining the passing cost of the vehicle to be detected passing through each gate, wherein the passing cost of the vehicle to be detected passing through each gate is a reference value corresponding to the passing standard time interval in which the passing time of the vehicle to be detected passing through each gate is located; determining the stability degree of the passing cost of the vehicle to be detected in the travel according to the mathematical variance of the passing cost of the vehicle to be detected passing through each gate; and determining whether the vehicle to be detected in the travel has abnormal vehicle speed according to the stability degree of the passing cost of the vehicle to be detected in the travel.

12. The abnormal vehicle identifying apparatus according to claim 11, wherein the abnormal vehicle identifying apparatus is configured to perform an operation of implementing the abnormal vehicle identifying method according to any one of claims 1 to 10.

13. An abnormal vehicle recognition device characterized by comprising:

a memory to store instructions;

a processor configured to execute the instructions to cause the abnormal vehicle identification apparatus to perform an operation of implementing the abnormal vehicle identification method according to any one of claims 1 to 10.

14. A computer-readable storage medium storing computer instructions which, when executed by a processor, implement the abnormal vehicle identification method of any one of claims 1-10.