CN113269060B

CN113269060B - Vehicle illegal behavior review method and device, electronic equipment and storage medium

Info

Publication number: CN113269060B
Application number: CN202110514020.0A
Authority: CN
Inventors: 史晓蒙; 魏健康; 韦创; 张伟; 马洪民; 李元齐
Original assignee: China Hualu Group Co Ltd; Beijing E Hualu Information Technology Co Ltd
Current assignee: China Hualu Group Co Ltd; Beijing E Hualu Information Technology Co Ltd
Priority date: 2021-05-11
Filing date: 2021-05-11
Publication date: 2024-02-13
Anticipated expiration: 2041-05-11
Also published as: CN113269060A

Abstract

The invention provides a vehicle illegal behavior review method, a device, electronic equipment and a storage medium, wherein the method comprises the following steps: obtaining vehicle illegal behavior information to be audited, wherein the vehicle illegal behavior information comprises illegal types; determining a corresponding illegal behavior review algorithm according to the illegal category; and determining the review result of the illegal behaviors according to the illegal behavior review algorithm and the vehicle illegal behavior information. By implementing the invention, the corresponding review algorithm can be automatically invoked to realize automatic review, so that the review efficiency is improved, the manual review load is reduced, the labor cost is reduced, the subjective factors during manual review can be avoided, and the review accuracy is improved.

Description

Vehicle illegal behavior review method and device, electronic equipment and storage medium

Technical Field

The invention relates to the field of computer image processing, in particular to a vehicle illegal behavior review method, a device, electronic equipment and a storage medium.

Background

With the increase of urban vehicles, front-end equipment continuously updates, and captured illegal action data continuously increases, but in the illegal action data, as a camera changes along with time, the complexity of a road environment increases, and a large amount of data of non-motor vehicles, pedestrians and special vehicles are misshot, so that a large amount of misjudgment data is generated. In the related art, in order to remove misjudgment data, a large amount of manual checking is needed, the work task is heavy, a large amount of repeated work exists, and meanwhile, the phenomenon of review errors is extremely easy to occur.

Disclosure of Invention

In view of the above, the embodiment of the invention provides a vehicle illegal behavior review method, a device, an electronic device and a storage medium, which are used for solving the defects that in the prior art, a large amount of manpower is required to be provided for manual review in order to remove misjudgment data, a large amount of repeated work exists, and review errors are extremely easy to occur.

According to a first aspect, an embodiment of the present invention provides a vehicle illegal activity review method, including the steps of: obtaining vehicle illegal behavior information to be audited, wherein the vehicle illegal behavior information comprises illegal types; determining a corresponding illegal behavior review algorithm according to the illegal category; and determining the review result of the illegal behaviors according to the illegal behavior review algorithm and the vehicle illegal behavior information.

Optionally, the vehicle illegal action information includes an image acquisition device number for acquiring the vehicle illegal action, a plurality of pieces of vehicle driving image information to be audited and illegal vehicle license plate number information, and the determining, according to the illegal action review algorithm and the vehicle illegal action information, a review result of the illegal action includes: extracting license plate number information in the vehicle driving image information; when the extracted license plate number information is consistent with the illegal license plate number information, determining lane line information of the illegal act according to the number of the image acquisition equipment; and carrying out illegal action judgment on the lane line information and the driving image information of the plurality of vehicles to be audited according to the illegal action review algorithm to obtain a review result of the illegal action.

Optionally, the type of the violation is not running according to a guide line, and the performing, according to the rule-breaking behavior review algorithm, rule-breaking behavior judgment on the lane line information and the driving image information of the plurality of vehicles to be reviewed, to obtain a rule-breaking behavior review result, includes: determining a vehicle driving path according to the plurality of pieces of vehicle driving image information to be checked; when the vehicle driving path has an intersection point with a rule line in the lane line information, the review result is that the vehicle is effective to run without being driven according to a guide line; or (b)

When the illegal category is line pressing running, the method for judging the illegal behavior of the lane line information and the driving image information of a plurality of vehicles to be checked according to the illegal behavior review algorithm, to obtain a review result of the illegal behavior, comprises the following steps: determining the relation between the diagonal coordinates of the vehicle and the lane lines according to the vehicle driving image information to be audited; and when the diagonal coordinates of the vehicle are at the two ends of the lane line, the rechecking result shows that the vehicle is effective in line pressing running.

Optionally, when the type of the illegal activity is reverse driving, the performing, according to the illegal activity review algorithm, the illegal activity judgment on the lane line information and the driving image information of the plurality of vehicles to be reviewed to obtain a review result of the illegal activity, including: determining a vehicle driving direction according to the plurality of pieces of vehicle driving image information to be checked; and when the driving direction of the vehicle is opposite to the direction of the lane line in the lane information, the review result is that the vehicle is effective to run reversely.

Optionally, the vehicle illegal action information includes an image acquisition device number for acquiring the vehicle illegal action and a plurality of pieces of vehicle driving image information to be checked, and the determining the checking result of the illegal action according to the illegal action checking algorithm and the vehicle illegal action information includes: determining lane attributes of the illegal behaviors according to the numbers of the image acquisition equipment, wherein the lane attributes comprise a target vehicle type traffic limiting lane, a parking forbidden lane and a single lane; when the lane attribute of the illegal act is inconsistent with the lane attribute in the vehicle illegal act information to be checked, the review result is invalid.

Optionally, the vehicle illegal action information further includes illegal vehicle license plate number information, and the lane attribute is a target vehicle type traffic limiting lane, and further includes: when the lane attribute of the illegal act is consistent with the lane attribute in the vehicle illegal act information to be audited, extracting license plate number information and vehicle type information in the vehicle driving image information; when the license plate number information is consistent with the license plate number information of the illegal vehicle and the vehicle type information is consistent with the target vehicle type, the review result is that the illegal behavior is effective.

Optionally, the vehicle illegal action information further includes illegal vehicle license plate number information, and the lane attribute is a parking forbidden lane, and further includes: when the lane attribute of the illegal act is consistent with the lane attribute in the vehicle illegal act information to be audited, extracting license plate number information in the vehicle driving image information; when the license plate number information is consistent with the illegal vehicle license plate number information, judging whether vehicles corresponding to the license plate number information are in a forbidden parking area in a plurality of pieces of vehicle driving image information; when the vehicles corresponding to the license plate number information are in the forbidden parking areas in the plurality of pieces of vehicle driving image information, the review result is that the illegal behaviors are effective.

Optionally, the vehicle illegal action information further includes illegal vehicle license plate number information, and the lane attribute is a single lane, and further includes: when the lane attribute of the illegal act is consistent with the lane attribute in the vehicle illegal act information to be audited, extracting license plate number information in the vehicle driving image information; when the license plate number information is consistent with the illegal vehicle license plate number information, positioning a target illegal vehicle; determining the vehicle type of the target illegal vehicle according to the target model; when the vehicle type is not the target vehicle type, the review result is that the illegal action is valid.

Optionally, the vehicle illegal action information includes a plurality of pieces of vehicle driving image information to be checked and illegal vehicle license plate number information, when the type of the illegal action is red light running, determining a review result of the illegal action according to the illegal action review algorithm and the vehicle illegal action information, including: extracting license plate number information in the vehicle driving image information; when the extracted license plate number information is consistent with the illegal vehicle license plate number information, judging whether the traffic light is a red light currently according to the vehicle driving image information; when the traffic light is currently a red light, judging whether the vehicle position corresponding to the license plate number information exceeds a detection area or not; and when the vehicle position corresponding to the license plate number information exceeds the detection area, the review result is that the illegal act is effective.

Optionally, the vehicle illegal action information includes a plurality of pieces of vehicle driving image information to be checked and illegal vehicle license plate number information, and determining a review result of the illegal action according to the illegal action review algorithm and the vehicle illegal action information includes: extracting license plate number information in the vehicle driving image information; when the extracted license plate number information is consistent with the illegal vehicle license plate number information, vehicle detection is carried out according to a target vehicle positioning model, and a vehicle image corresponding to the illegal vehicle license plate number information is obtained; according to the target personnel positioning model, detecting the personnel position of the vehicle image to obtain a driver image; inputting the driver images into a personnel behavior classification model to obtain classification results; and determining a review result of the illegal act according to the classification result.

According to a second aspect, an embodiment of the present invention provides a vehicle illicit-action review device, including: the information acquisition module is used for acquiring the vehicle illegal action information to be audited, wherein the vehicle illegal action information comprises illegal types; the review algorithm determining module is used for determining a corresponding illegal behavior review algorithm according to the illegal category; and the review result determining module is used for determining the review result of the illegal behaviors according to the illegal behavior review algorithm and the vehicle illegal behavior information.

According to a third aspect, an embodiment of the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the steps of the vehicle illicit-behavior review method according to the first aspect or any implementation manner of the first aspect when the processor executes the program.

According to a fourth aspect, an embodiment of the present invention provides a storage medium having stored thereon computer instructions, which when executed by a processor, implement the steps of the vehicle illicit-behavior review method of the first aspect or any implementation of the first aspect.

The technical scheme of the invention has the following advantages:

According to the vehicle illegal behavior review method/device provided by the embodiment of the invention, the corresponding illegal behavior review algorithm is determined according to the illegal types in the to-be-reviewed vehicle illegal behavior information, the review result of the illegal behavior is determined according to the corresponding illegal behavior review algorithm, and the corresponding review algorithm can be automatically invoked to realize automatic review aiming at a large amount of misjudgment data, so that the review efficiency is improved, the manual review load is reduced, the labor cost is reduced, the subjective factors during manual review can be avoided, and the review accuracy is improved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flowchart of a specific example of a vehicle illicit review method in an embodiment of the invention;

FIG. 2 is a schematic block diagram of a specific example of a vehicle illicit-action review device in an embodiment of the invention;

Fig. 3 is a schematic block diagram of a specific example of an electronic device in an embodiment of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made apparent and fully in view of the accompanying drawings, in which some, but not all embodiments of the invention are shown. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

In the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings, are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.

In the description of the present invention, it should be noted that, unless explicitly specified and limited otherwise, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be either fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; the two components can be directly connected or indirectly connected through an intermediate medium, or can be communicated inside the two components, or can be connected wirelessly or in a wired way. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.

In addition, the technical features of the different embodiments of the present invention described below may be combined with each other as long as they do not collide with each other.

The embodiment provides a vehicle illegal action review method, as shown in fig. 1, comprising the following steps:

s101, obtaining vehicle illegal action information to be audited, wherein the vehicle illegal action information comprises illegal categories;

for example, the vehicle infraction information to be audited may be obtained by the front-end camera, where the vehicle infraction information may include a device number of the front-end camera, license plate number information, infraction time, infraction place, at least one piece of infraction picture information, infraction type, vehicle coordinate, license plate coordinate, etc., where the infraction type represents the initially determined infraction type, including reverse driving, driving without guidance, red light running, line pressing driving, stopping, distraction driving, unbelted belt, illegal driving, etc., different infraction types may be distinguished by different numbers, for example, reverse driving number 1301, driving without guidance No. 1208, line pressing No. 1345, red light running No. 1625, unbelted driving No. 6011, distraction driving No. 1223, driving target vehicle type No. 1344 on a limited lane, bidirectional driving No. 7074, stopping vehicle No. 13441 or 13451, etc. The kind of the illegal behavior information of the vehicle and the kind of the illegal behavior are not limited in this embodiment, and can be determined by those skilled in the art according to the need.

The specific obtained information of the vehicle illegal behaviors to be audited can comprise:

{ "isolation_xh": 1201120001528031, # device number

＂Xh＂：＂＂，

＂VioType＂：＂Violation＂，

"Hphm": "jin FAN 788", # license plate number

＂Hpzl＂：＂02＂，

"Wfsj": "2020-11-26 11:38: 59', time of violation #

"Wfdd": "201050017320" # illegal location number

"Wfdz": ' Jinqi road and prosperous road intersection west east ', ' violation address

Code for # illicit behavior

＂Wfxw＂：＂1208＂，

＂Clsd＂：0，

＂Clxs＂：0，

＂Sjly＂：＂1＂，

＂Sfqj＂：＂00＂，

＂Qjys＂：0，

＂Tztp1＂：＂＂，

＂Tztp2＂：＂＂，

Picture for law breaking #

1＂Zjwj1＂：＂http：//172.30.2.9/weifa/2020/11/26/201050017320/baf906ae541945909febe6e6c8ccee4b.jpg＂，

Picture for law breaking #

2＂Zjwj2＂：＂http：//172.30.2.9/weifa/2020/11/26/201050017320/e0a118a22c6943899e0afb6f7ba41d01.jpg＂，

Picture for law breaking #

3＂Zjwj3＂：＂http：//172.30.2.9/weifa/2020/11/26/201050017320/1ccfc21b165042ba83a2d51691fb15ac.jpg＂，

Picture 4 for law violation

＂Zjwj4＂：＂＂，

＂Shbj＂：＂00＂，

＂Tblhy＂：0，

＂Cjjg＂：＂121200000000＂，

＂Scbj＂：0，

＂Cdbh＂：1，

＂Cjsj＂：＂2020－11－26 11：38：59＂，

＂Csbl＂：0，

＂Fxbh＂：＂02＂，

＂Cscs＂：0，

＂Qjjl＂：0，

＂Zdxs＂：0，

＂Sbbh＂：＂121200000000099998＂，

＂Bdbj＂：0，

＂Sbbj＂：1，

＂Splj＂：＂http：//10.112.93.2：7777/hikvision/ch6/starttime＝20201126T113855Z/endtime＝20201126T113905Z/1.mp4＂，

＂AiShbj＂：＂00＂，

＂AiClbj＂：＂00＂

Vehicle coordinate 1

“Clzb1”：”＊＊，＊＊，＊＊，＊＊”

Vehicle coordinate 2

“Clzb2”：”＊＊，＊＊，＊＊，＊＊”

Vehicle coordinate 3

“Clzb3”：”＊＊，＊＊，＊＊，＊＊”

Number plate coordinate 1

“CPzb1”：”＊＊，＊＊，＊＊，＊＊”

Number plate coordinate 2

“CPzb2”：”＊＊，＊＊，＊＊，＊＊”

# license plate coordinate 3

“CPzb3”：”＊＊，＊＊，＊＊，＊＊”}

S102, determining a corresponding illegal behavior review algorithm according to the type of the illegal;

the manner of determining the corresponding illicit review algorithm may be, for example, automatically invoking the corresponding pre-stored illicit review algorithm when the illicit category in the vehicle illicit information is detected, depending on the illicit category. For example, when the illegal category number 1301 is detected, a review algorithm corresponding to the retrograde is invoked. The method for determining the corresponding illicit action review algorithm according to the illicit category is not limited in this embodiment, and can be determined by a person skilled in the art according to the need.

S103, determining the review result of the illegal behaviors according to the review algorithm of the illegal behaviors and the information of the illegal behaviors of the vehicle.

Illustratively, taking the case of the type of the violation as the reverse direction (the number of the detected type of the violation is 1301), a review algorithm corresponding to the reverse direction may be to extract the number of the image acquisition device therein, and the drawing configuration information of the monitoring area corresponding to the number of the image acquisition device is stored in the system database in advance. The database information can be matched through the number of the image acquisition equipment, and the line drawing configuration information corresponding to the monitoring area of the number of the current image acquisition equipment is obtained, so that the lane line direction and the lane detection area information are obtained.

Then extracting license plate coordinates in the vehicle illegal action information, intercepting a license plate coordinate local picture to carry out license plate number recognition, and checking whether license plate characters in the picture are identical with license plate number results in the vehicle illegal action information or not; if the reverse running result is different, returning to be invalid; if the vehicle running direction is the same, the vehicle running direction is calculated according to license plate coordinates corresponding to a plurality of illegal events, whether the vehicle running direction is opposite to the direction of the lane lines is judged, whether the vehicle running direction is illegal or not is further judged, if so, the reverse running review result is effective, and if the vehicle running direction is the same, the reverse running review result is ineffective. The review algorithm is different for different illegal behaviors, and the embodiment is not limited to the above, and can be determined by a person skilled in the art according to the need. Experiments prove that the accuracy of the review result can reach 95% or more for different illegal types.

The review result output format of the illegal act can be expressed as:

json{

result of determination # valid (indeed illegal event): 1, invalid: 0

jsonType：’AI’}

According to the vehicle illegal behavior review method provided by the embodiment, corresponding illegal behavior review algorithms are determined according to the illegal types in the to-be-reviewed vehicle illegal behavior information, the review results of the illegal behaviors are determined according to the corresponding illegal behavior review algorithms, and the corresponding review algorithms can be automatically invoked to realize automatic review aiming at a large amount of misjudgment data, so that the review efficiency is improved, the manual review load is reduced, the labor cost is reduced, subjective factors during manual review can be avoided, and the review accuracy is improved.

As an optional implementation manner of this embodiment, the vehicle illegal action information includes an image acquisition device number for acquiring the vehicle illegal action, a plurality of pieces of vehicle driving image information to be checked, and illegal vehicle license plate number information, and determining a review result of the illegal action according to a review algorithm of the illegal action and the vehicle illegal action information, including:

firstly, extracting license plate number information in vehicle driving image information;

the method of extracting the license plate number information in the vehicle driving image information may be to input the vehicle driving image information into a pre-trained license plate number recognition neural network, or may be to pre-process the image, and adopt license plate positioning, segmentation, character segmentation and character recognition modes. Through the license plate recognition mode, the accuracy rate of license plate recognition can be higher than 98%. The manner of extracting the license plate number information from the vehicle driving image information is not limited in this embodiment, and can be determined as required by those skilled in the art.

Secondly, when the extracted license plate number information is consistent with the license plate number information of the illegal vehicle, determining lane line information of illegal behaviors according to the number of the image acquisition equipment;

for example, the image capturing device is numbered information which is matched with each piece of monitoring road section information one by one, for example, the image capturing device is numbered 1201120001528031, and then the corresponding monitoring road section information in the system is road information of XX street XX and XX, including line drawing configuration information of lane attribute, lane line direction, lane line type and the like. Because the image acquisition device may not directly acquire the lane line information from the image due to the problem of the shielding object or the shooting angle, in this embodiment, according to the number of the image acquisition device, the manner of determining the lane line information of the illegal act may be to match the monitored road section information corresponding to the number of the image acquisition device in the system, so as to extract the lane line information from the monitored road section information.

And thirdly, carrying out illegal action judgment on the lane line information and the driving image information of the plurality of vehicles to be audited according to an illegal action rechecking algorithm to obtain a rechecking result of the illegal action.

For example, in this embodiment, the case where the type of the violation is not traveling along the guide line (the type of the violation is detected as 1208), and the rule-breaking behavior is determined according to the rule-breaking behavior review algorithm on the lane line information and the plurality of pieces of vehicle driving image information to be reviewed, so as to obtain the review result of the rule-breaking behavior, which may be to determine the vehicle driving path according to the plurality of pieces of vehicle driving image information to be reviewed, specifically including: and compared with the mode of determining the vehicle driving path based on the video segment, the method can judge illegal time through a plurality of pictures, reduces storage space and reduces processing capacity.

When the vehicle driving path has an intersection point with a rule line in the lane line information, the review result is that the vehicle is effective to run without being driven according to a guide line; when the driving path of the vehicle has no intersection point with the regular line in the lane line information, the review result is that the vehicle is not invalid to run according to the guide line. The review algorithm is different for different kinds of violations, and the embodiment is not limited to this, and can be determined by those skilled in the art according to requirements.

According to the vehicle illegal action review method provided by the embodiment, the license plate number information in the image information is firstly extracted, and when the license plate number information is consistent with the illegal vehicle license plate number information, the subsequent review is executed, so that the problem that the accurate lane line information cannot be directly obtained from the image due to the fact that the image acquisition equipment is offset in shielding objects or shooting angles is avoided, the excessive judgment process is executed when the license plate number information is inconsistent with the illegal vehicle license plate number information, the calculation amount of review is reduced, in addition, lane line information of illegal actions is determined according to the number of the image acquisition equipment, and the accuracy of vehicle illegal action review is improved.

As an optional implementation manner of this embodiment, when the illegal category is line-pressing running, performing illegal behavior judgment on lane line information and a plurality of pieces of vehicle driving image information to be audited according to an illegal behavior review algorithm, to obtain a review result of the illegal behavior, including: determining the relation between the diagonal coordinates of the vehicle and the lane lines according to the driving image information of the vehicle to be audited; when the diagonal coordinates of the vehicle are at the two ends of the lane line, the rechecking result shows that the vehicle is effective in line pressing running.

For example, when the type of the violation is line-pressing running (the type of the violation is detected as 1345), the method of determining the vehicle driving path according to the plurality of pieces of vehicle driving image information to be checked may be to extract the vehicle coordinate information in the vehicle driving image information to be checked, determine whether the diagonal coordinates of the vehicle are at the two ends of the lane line, and when the diagonal coordinates of the vehicle are at the two ends of the lane line, the rechecking result is that the line-pressing running of the vehicle is effective; and when the diagonal coordinates of the vehicle are not at the two ends of the lane line, the rechecking result shows that the line pressing running of the vehicle is invalid.

As an optional implementation manner of this embodiment, when the type of the violation is reverse driving (the type number of the violation is 1301 is detected), the rule-breaking behavior determination is performed on the lane line information and the plurality of pieces of vehicle driving image information to be checked according to the rule-breaking behavior review algorithm, so as to obtain a rule-breaking behavior review result, including: determining a vehicle driving direction according to the plurality of pieces of vehicle driving image information to be checked; when the driving direction of the vehicle is opposite to the direction of the lane line in the lane information, the review result is that the vehicle is effective to run reversely.

For example, the method of determining the driving direction of the vehicle according to the plurality of pieces of vehicle driving image information to be checked may be to calculate license plate coordinates or vehicle coordinates in the plurality of pieces of vehicle driving image information to be checked according to the time sequence of acquiring the vehicle driving images, so as to obtain the driving direction of the vehicle. When the driving direction of the vehicle is opposite to the direction of the lane line in the lane information, the review result is that the vehicle is effective in reverse driving; when the driving direction of the vehicle is the same as the lane line direction in the lane information, the review result is that the vehicle is invalid to run in the reverse direction.

As an optional implementation manner of this embodiment, the vehicle illegal activity information includes an image acquisition device number for acquiring the vehicle illegal activity and a plurality of pieces of vehicle driving image information to be checked, and determining, according to an illegal activity review algorithm and the vehicle illegal activity information, a review result of the illegal activity includes: determining lane attributes of illegal behaviors according to the number of the image acquisition equipment, wherein the lane attributes comprise a target vehicle type traffic limiting lane, a parking forbidding lane and a single lane; when the lane attribute of the illegal act is inconsistent with the lane attribute in the information of the illegal act of the vehicle to be checked, the re-checking result is invalid.

Illustratively, the system database stores the line drawing configuration information of the monitoring area corresponding to the image acquisition device number in advance. According to the number of the image acquisition equipment, the mode of determining the lane attribute with illegal actions can be that the number of the image acquisition equipment is matched with the database information to obtain the line drawing configuration information corresponding to the monitoring area of the number of the current image acquisition equipment, so as to obtain the lane attribute. The lane attributes comprise a target vehicle type traffic limiting lane, a parking forbidding lane, a single-lane and the like, wherein the target vehicle type traffic limiting lane can represent the lane or the road section forbids the target vehicle type from running, such as a large truck forbid driving lane/road section; the parking-forbidden road can characterize the lane or the road section or a specific position for stopping; a single-lane may characterize the lane or the road segment as belonging to a single-lane within a certain period of time, or the lane is a single-lane throughout the period of time.

When the lane attribute of the illegal act, which is determined according to the number of the image acquisition equipment, is inconsistent with the lane attribute in the information of the illegal act of the vehicle to be audited, the illegal act recorded in the information of the illegal act of the vehicle to be audited does not occur, so that the review result is invalid. For example, when the lane attribute in the vehicle illegal behavior information to be checked is a large truck prohibited driving lane, but the lane attribute of the illegal behavior is determined according to the number of the image acquisition equipment and is not the large truck prohibited driving lane, the initial illegal behavior is considered to be misjudged, and the re-checking result is that the large truck on-road illegal behavior is invalid.

As an optional implementation manner of this embodiment, the vehicle illegal action information further includes license plate number information of an illegal vehicle, and the lane attribute is a target vehicle type traffic limiting lane, and further includes: when the lane attribute of the illegal act is consistent with the lane attribute in the vehicle illegal act information to be audited, extracting license plate number information and vehicle type information in the vehicle driving image information; when the license plate number information is consistent with the license plate number information of the illegal vehicle and the vehicle type information is consistent with the target vehicle type, the review result is that the illegal behavior is effective.

For example, when the detected type of the violation is the target vehicle type on the limited traffic lane, that is, the number of the violation is 1344, license plate number information and vehicle type information in the vehicle driving image information are extracted, the vehicle driving image information is subjected to screenshot according to license plate coordinates and vehicle coordinates, and the license plate image or the vehicle image after the screenshot is input into a pre-trained license plate number recognition model or vehicle type recognition model, so that the license plate number information and the vehicle type information are obtained. The method for extracting the license plate number information and the vehicle type information in the vehicle driving image information is not limited in this embodiment, and can be determined by a person skilled in the art according to needs.

When the license plate number information is inconsistent with the illegal license plate number information in the illegal behavior information of the vehicle to be checked, the illegal behavior information of the vehicle to be checked is inaccurate, and the re-checking result is that the illegal behavior is invalid; when the license plate number information is consistent with the illegal license plate number information in the illegal behavior information of the vehicle to be checked, but the vehicle type information is inconsistent with the target vehicle type, the license plate number information indicates that the vehicle corresponding to the license plate number information does not belong to the illegal behavior when running on the limited traffic lane of the target vehicle type, and the review result is that the illegal behavior is invalid; when the license plate number information is consistent with the illegal license plate number information in the illegal behavior information of the vehicle to be checked, and the vehicle type information is consistent with the target vehicle type, the fact that the target vehicle type corresponding to the license plate number information runs on the limited traffic lane of the target vehicle type is truly illegal behavior is indicated, and the re-checking result is that the illegal behavior is effective.

For example, when reviewing, firstly, a plurality of license plate coordinates of a plurality of pieces of vehicle driving image information to be reviewed are extracted, the plurality of pieces of vehicle driving image information to be reviewed can be three, a license plate coordinate local picture is intercepted, license plate number recognition is carried out, and whether license plate characters in the picture are identical with front-end equipment recognition results or not is checked; if the two types of the information are different, returning to pre-examination and invalidation; if the vehicle coordinates are the same, the vehicle coordinates are read, the vehicle screenshot is extracted, whether the vehicle type is a large truck is judged, if yes, the pre-examination is valid, and if not, the pre-examination is invalid.

As an optional implementation manner of this embodiment, the vehicle illegal action information further includes license plate number information of an illegal vehicle, and the lane attribute is a parking forbidden lane, and further includes:

firstly, when detecting that the illegal type is parking of a vehicle in a parking forbidden lane, namely, the illegal type number is 13441 or 13451 (the forbidden areas corresponding to the two numbers are different in representation form, for example, 13441 represents that the forbidden area is a specific area of a yellow cross grid, and 13451 represents that the forbidden area is a whole lane), judging whether the lane attribute of illegal behaviors is consistent with the lane attribute in the vehicle illegal behavior information to be audited, and extracting license plate number information in the driving image information of a plurality of vehicles; the manner of extracting the license plate number information in the vehicle driving image information is described in the above embodiment, and the content of the license plate number information is not described herein.

Secondly, when the license plate number information is consistent with the license plate number information of the illegal vehicle, judging whether the vehicle corresponding to the license plate number information is in a forbidden parking area in the plurality of vehicle driving image information; when the vehicles corresponding to the license plate number information are in the forbidden parking areas in the plurality of pieces of vehicle driving image information, the review result is that the illegal behaviors are effective.

The parking prohibition area may be characterized as a parking prohibition area, or may be characterized as a parking prohibition lane, which is not limited in this embodiment, and may be determined by a person skilled in the art according to needs. The plurality of pieces of vehicle driving image information may be image information photographed at different times, the number of pieces of vehicle driving image information may be determined according to photographed time intervals and a parking time threshold value, and if the parking time threshold value indicates that the parking duration exceeds the threshold value, it may be determined that illegal parking is performed. For example, when the parking time threshold is 1 minute and the time interval of photographing is 20S one, then the number of pieces of vehicle driving image information may be 3. When the vehicles corresponding to the license plate number information are in the forbidden parking areas in the plurality of pieces of vehicle driving image information, the review result is that the illegal behaviors are effective; when the vehicles corresponding to the license plate number information are not in the forbidden parking areas in the plurality of pieces of vehicle driving image information, the review result is invalid in illegal actions.

As an optional implementation manner of this embodiment, the vehicle illegal action information further includes illegal vehicle license plate number information, and the lane attribute is a single lane, and further includes:

firstly, when the illegal type is detected to be the bidirectional running of a single lane, namely the illegal type number is 7074, judging whether the lane attribute of illegal behavior is consistent with the lane attribute in the illegal behavior information of the vehicle to be checked, extracting license plate number information in the driving image information of the vehicle, judging whether the license plate number information is consistent with the license plate number information of the illegal vehicle, and positioning the target illegal vehicle according to the license plate number information in the illegal behavior information of the vehicle to be checked when the license plate number information is consistent with the license plate number information of the illegal vehicle.

For example, the one-way lane representation comprises that the lane or the road section belongs to a one-way lane within a certain period of time, or that the lane full period is a one-way road. When the single-lane represents that the whole period of the lane is a single-lane, the mode of judging whether the lane attribute of the illegal behavior is consistent with the lane attribute in the vehicle illegal behavior information to be audited can be to match the corresponding lane attribute information in the database according to the number of the image acquisition equipment and judge whether the lane attribute information is a single-lane. When the single-lane represents the lane or the road section belongs to the single-lane in a certain time period, the method for judging the lane attribute of the illegal act further comprises judging whether the time of the illegal act is within a single-lane limiting time range, such as (7:00-19:00), if not, the review result is invalid.

When the lane attribute of the illegal act is consistent with the lane attribute in the information of the illegal act of the vehicle to be audited, namely the lane attribute of the illegal act actually belongs to a single lane, extracting license plate information in the vehicle driving image information, judging whether the license plate information is consistent with the license plate information of the illegal vehicle in the information of the illegal act of the vehicle to be audited, and when the detected license plate information is inconsistent with the license plate information of the illegal vehicle in the information of the illegal act of the vehicle to be audited, the review result is invalid; when the detected license plate number information of the vehicle is consistent with the license plate number information of the illegal vehicle in the information of the illegal behavior of the vehicle to be checked, the target illegal vehicle is positioned in the vehicle driving image information according to the license plate number information in the information of the illegal behavior of the vehicle to be checked, and the method for positioning the target illegal vehicle can be to identify each vehicle position in the picture through a vehicle structuring algorithm. The vehicle driving image may be composed of a plurality of images, for example, 3 images, and the largest/nearest vehicle driving image of the target is preferentially selected, if the first image does not identify the target illegal vehicle, the 2 nd image is taken, and the like, until the target illegal vehicle is identified.

Secondly, determining the vehicle type of the target illegal vehicle according to the target model;

by way of example, the target model may be a ResNet50 motor vehicle attribute model, and the accuracy of vehicle type identification of target offending vehicles may be up to 98% and above. From the target model, a vehicle type of the target offending vehicle may be determined. The present embodiment does not limit the object model, and those skilled in the art can determine the object model according to the need.

Again, when the vehicle type is not the target vehicle type, then the review result is that the illegal action is valid.

For example, the target vehicle type may refer to a vehicle that is not defined by a single lane rule, such as a bus. When the vehicle type is not the bus, the review result is that the illegal act is effective, so that misjudgment on the vehicle which is not limited by the rule of the single lane when the single lane illegal travel review is carried out is avoided, and the accuracy of the review is improved.

As an optional implementation manner of this embodiment, the vehicle illegal action information includes a plurality of pieces of vehicle driving image information to be checked and illegal vehicle license plate number information, and when the type of the illegal action is red light running (when the type of the illegal action is detected as 1625), determining a review result of the illegal action according to an illegal action review algorithm and the vehicle illegal action information, including:

Firstly, extracting license plate number information in vehicle driving image information; the details are referred to the corresponding parts of the above embodiments, and are not repeated here.

Secondly, when the extracted license plate number information is consistent with the illegal vehicle license plate number information, judging whether the traffic light is a red light currently according to the vehicle driving image information; and when the traffic light is currently a red light, judging whether the vehicle position corresponding to the license plate number information exceeds a detection area. And when the vehicle position corresponding to the license plate number information exceeds the detection area, the review result is that the illegal act is effective.

When the extracted license plate number information is consistent with the illegal vehicle license plate number information, judging whether the traffic light is a red light currently according to the vehicle driving image information; when the extracted license plate number information is inconsistent with the illegal vehicle license plate number information, the review result is invalid in illegal behaviors. When the traffic light is a red light currently, judging whether the vehicle position corresponding to the license plate number information exceeds a detection area, and when the vehicle position corresponding to the license plate number information exceeds the detection area, the review result is that the illegal act is effective; when the red-green lamp is not the red lamp currently, the review result is that the illegal action is invalid.

As an optional implementation manner of this embodiment, the vehicle illegal action information includes a plurality of pieces of vehicle driving image information to be checked and illegal vehicle license plate number information, and when the type of the illegal action is distracted driving or unbelted driving (the type of the illegal action is detected as 1223 or the type of the illegal action is detected as 6011), the method for determining the review result of the illegal action according to the review algorithm of the illegal action and the vehicle illegal action information includes:

firstly, extracting license plate number information in vehicle driving image information; when the extracted license plate number information is consistent with the illegal vehicle license plate number information, vehicle detection is carried out according to the target vehicle positioning model, and a vehicle image corresponding to the illegal vehicle license plate number information is obtained;

for example, the manner of extracting the license plate number information in the vehicle driving image information is referred to the corresponding parts of the above embodiments, and will not be described herein. In general, the vehicle driving image to be checked is formed by splicing pictures of two frames in sequence, so that half of the vehicle driving image is needed to be cut out for license plate recognition, and when the license plate recognition result is inconsistent with the license plate information of the illegal vehicle, the re-checking result of the illegal behavior is invalid. When the extracted license plate number information is consistent with the illegal vehicle license plate number information, vehicle detection is carried out according to the target vehicle positioning model, and a vehicle image corresponding to the illegal vehicle license plate number information is obtained. The target vehicle localization model may be YOLO V4, among others.

Secondly, according to the target personnel positioning model, carrying out personnel position detection on the vehicle image to obtain a driver image;

for example, when a vehicle image is obtained, the vehicle image may be input to a target person positioning model, which may be a YOLO V3 window model, an accurate window position is detected using the YOLO 3 window model, and a right half area is cut out, resulting in a driver image.

Then, inputting the driver image into a personnel behavior classification model to obtain a classification result; and determining the review result of the illegal act according to the classification result.

For example, when the type of the illegal act is the distraction driving, the human behavior classification model may be the distraction driving classification model, when the driver is detected to be making a call, the obtained classification result is that the driver is driving distraction, then the review result of the illegal act is valid, otherwise, it is invalid.

When the illegal types are unbelted, the personnel behavior classification model can be an unbelted classification model, and when the fact that the driver is unbelted is detected, the obtained classification result is that the driver drives without belting, and then the review result of the illegal behaviors is valid, otherwise, the review result is invalid.

The embodiment provides a vehicle illegal action review device, as shown in fig. 2, including:

the information acquisition module 201 is configured to acquire vehicle illegal activity information to be audited, where the vehicle illegal activity information includes an illegal category; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The review algorithm determining module 202 is configured to determine a corresponding review algorithm of the illegal act according to the type of the illegal act; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

And the review result determining module 203 is configured to determine a review result of the illegal act according to the illegal act review algorithm and the vehicle illegal act information. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

As an alternative implementation manner of this embodiment, the review result determining module 203 includes:

the first license plate number extraction module is used for extracting license plate number information in the vehicle driving image information; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The lane line determining module is used for determining lane line information of the illegal act according to the number of the image acquisition equipment when the extracted license plate number information is consistent with the license plate number information of the illegal vehicle; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

And the first review result determination submodule is used for judging the illegal behaviors according to the lane line information and the driving image information of the vehicles to be reviewed according to the illegal behaviors review algorithm to obtain the review result of the illegal behaviors. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

the path determining module is used for determining a vehicle driving path according to the plurality of pieces of vehicle driving image information to be audited; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The non-guide line running review module is used for enabling the review result to be that the vehicle is effective in non-guide line running when the vehicle driving path has an intersection point with a rule line in the lane line information; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein. Or (b)

The review result determining module 203 includes:

the vehicle diagonal coordinate relation determining module is used for determining the relation between the vehicle diagonal coordinates and the lane lines according to the vehicle driving image information to be audited; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

And the vehicle line pressing review module is used for enabling the review result to be that the vehicle line pressing running is effective when the diagonal coordinates of the vehicle are at the two ends of the lane line. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

the driving direction determining module is used for determining the driving direction of the vehicle according to the driving image information of the plurality of vehicles to be audited; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

And the reverse review module is used for enabling the vehicle to run reversely and effectively when the driving direction of the vehicle is opposite to the direction of the lane line in the lane information. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

the lane attribute determining module is used for determining lane attributes of the illegal behaviors according to the number of the image acquisition equipment, wherein the lane attributes comprise a target vehicle type traffic limiting lane, a parking forbidden lane and a single lane; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

And the second review result determination submodule is used for determining that the review result is invalid when the lane attribute of the illegal act is inconsistent with the lane attribute in the vehicle illegal act information to be reviewed. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

As an optional implementation manner of this embodiment, the vehicle illegal behavior review device further includes:

the license plate and vehicle type extraction module is used for extracting license plate information and vehicle type information in the vehicle driving image information when the lane attribute of the illegal action is consistent with the lane attribute in the vehicle illegal action information to be audited; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

And the target vehicle on-road review module is used for enabling the review result to be effective in illegal actions when the license plate number information is consistent with the license plate number information of the illegal vehicle and the vehicle type information is consistent with the target vehicle type. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The second license plate number extraction module is used for extracting license plate number information in the vehicle driving image information when the lane attribute of the illegal action is consistent with the lane attribute in the vehicle illegal action information to be audited; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The vehicle position judging module is used for judging whether the vehicle corresponding to the license plate number information is in a forbidden parking area in a plurality of pieces of vehicle driving image information or not when the license plate number information is consistent with the illegal vehicle license plate number information; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

And the forbidden and stopped review module is used for judging that the review result is valid for illegal actions when the vehicles corresponding to the license plate number information are in forbidden and stopped areas in the plurality of pieces of vehicle driving image information. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

the third license plate number extraction module is used for extracting license plate number information in the vehicle driving image information when the lane attribute of the illegal action is consistent with the lane attribute in the vehicle illegal action information to be audited; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The target illegal vehicle positioning module is used for positioning a target illegal vehicle when the license plate number information is consistent with the illegal vehicle license plate number information; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The vehicle type determining module is used for determining the vehicle type of the target illegal vehicle according to the target model; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

And the target vehicle review module is used for enabling the review result to be valid for illegal behaviors when the vehicle type is not the target vehicle type. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

the fourth license plate number extraction module is used for extracting license plate number information in the vehicle driving image information; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The red light judging module is used for judging whether the traffic light is a red light currently according to the vehicle driving image information when the extracted license plate number information is consistent with the illegal vehicle license plate number information; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The area detection module is used for judging whether the vehicle position corresponding to the license plate number information exceeds a detection area or not when the traffic light is a red light currently; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

And the red light running review module is used for enabling the review result to be effective in illegal actions when the vehicle position corresponding to the license plate number information exceeds the detection area. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

a fifth license plate number extraction module for extracting license plate number information in the vehicle driving image information; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The vehicle image positioning module is used for carrying out vehicle detection according to a target vehicle positioning model when the extracted license plate number information is consistent with the illegal vehicle license plate number information, so as to obtain a vehicle image corresponding to the illegal vehicle license plate number information; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The target personnel positioning module is used for detecting the personnel position of the vehicle image according to the target personnel positioning model to obtain a driver image; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

The classification module is used for inputting the driver images into a personnel behavior classification model to obtain classification results; the specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

And the classification result review module is used for determining the review result of the illegal act according to the classification result. The specific content refers to the corresponding parts of the above method embodiments, and will not be described herein.

Embodiments of the present application also provide an electronic device, as shown in fig. 3, a processor 310 and a memory 320, where the processor 310 and the memory 320 may be connected by a bus or other means.

The processor 310 may be a central processing unit (Central Processing Unit, CPU). The processor 310 may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSPs), application specific integrated circuits (Application Specific Integrated Circuit, ASICs), field programmable gate arrays (Field-Programmable Gate Array, FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above.

The memory 320 is used as a non-transitory computer readable storage medium, and can be used to store non-transitory software programs, non-transitory computer executable programs, and modules, such as program instructions/modules corresponding to the vehicle illicit review method in the embodiments of the invention. The processor executes various functional applications of the processor and data processing by running non-transitory software programs, instructions, and modules stored in memory.

Memory 320 may include a storage program area that may store an operating system, at least one application program required for functionality, and a storage data area; the storage data area may store data created by the processor, etc. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, memory 320 may optionally include memory located remotely from the processor, which may be connected to the processor via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The one or more modules are stored in the memory 320, which when executed by the processor 310, perform a vehicle illicit review method as in the embodiment of fig. 1.

The details of the above electronic device may be understood correspondingly with respect to the corresponding related descriptions and effects in the embodiment shown in fig. 1, which are not repeated herein.

The present embodiment also provides a computer storage medium storing computer-executable instructions that can execute the vehicle illicit behavior review method in any of the above-described method embodiment 1. Wherein the storage medium may be a magnetic Disk, an optical Disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a Flash Memory (Flash Memory), a Hard Disk (HDD), or a Solid State Drive (SSD); the storage medium may also comprise a combination of memories of the kind described above.

It is apparent that the above examples are given by way of illustration only and are not limiting of the embodiments. Other variations or modifications of the above teachings will be apparent to those of ordinary skill in the art. It is not necessary here nor is it exhaustive of all embodiments. While still being apparent from variations or modifications that may be made by those skilled in the art are within the scope of the invention.

Claims

1. The vehicle illegal behavior review method is characterized by comprising the following steps of:

obtaining vehicle illegal action information to be audited, wherein the vehicle illegal action information comprises an illegal category, an image acquisition equipment number for obtaining the vehicle illegal action, a plurality of pieces of vehicle driving image information to be audited and illegal vehicle license plate number information, and the illegal category comprises: the method comprises the steps of driving in reverse, driving without a guide line, running red light, line pressing, stopping, distracted driving, unbuckling a safety belt and driving on a road in violation, wherein the image acquisition equipment number is used for acquiring lane line information or lane attributes of a pre-stored monitoring area corresponding to the image acquisition equipment number, and the lane attributes comprise a target vehicle type traffic limiting lane, a parking forbidden lane and a single-way lane;

Determining a prestored illegal behavior review algorithm corresponding to the illegal category according to the illegal category;

determining a review result of the illegal behaviors according to the illegal behavior review algorithm and the vehicle illegal behavior information;

extracting license plate number information in the vehicle driving image information; when the extracted license plate number information is consistent with the illegal vehicle license plate number information, if the illegal type is retrograde, not according to a guide line, line pressing, illegal stop or illegal road feeding, determining lane attribute or lane line information of the illegal behavior according to the number of the image acquisition equipment; carrying out illegal action judgment on the lane attribute or lane line information and the driving image information of a plurality of vehicles to be inspected according to the illegal action review algorithm to obtain a review result of the illegal action;

if the type of the violation is red light running, performing the violation judgment on the vehicle driving image information according to the violation review algorithm, wherein the method comprises the following steps: judging whether the traffic light is a red light currently according to the vehicle driving image information; when the traffic light is currently a red light, judging whether the vehicle position corresponding to the license plate number information exceeds a detection area or not; when the vehicle position corresponding to the license plate number information exceeds the detection area, the review result is that the illegal act is effective;

If the illegal category is distracted driving or unbelted driving, carrying out illegal action judgment on the vehicle driving image information and the illegal vehicle license plate number information according to the illegal action review algorithm, wherein the method comprises the following steps: and carrying out vehicle detection according to the target vehicle positioning model to obtain a vehicle image corresponding to the license plate number information of the illegal vehicle, carrying out personnel position detection on the vehicle image according to the target personnel positioning model to obtain a driver image, inputting the driver image into a personnel behavior classification model to obtain a classification result, and determining a review result of the illegal behavior according to the classification result.

2. The method according to claim 1, wherein if the type of the illegal act is not driving according to a guide line, performing illegal act judgment on the lane line information and a plurality of pieces of vehicle driving image information to be audited according to the illegal act review algorithm to obtain a review result of the illegal act, including:

determining a vehicle driving path according to the plurality of pieces of vehicle driving image information to be checked;

when the vehicle driving path has an intersection point with a rule line in the lane line information, the review result is that the vehicle is effective to run without being driven according to a guide line; or (b)

If the illegal category is line-pressing running, carrying out illegal action judgment on the lane line information and the plurality of pieces of vehicle driving image information to be checked according to the illegal action review algorithm to obtain a review result of the illegal action, wherein the method comprises the following steps of:

determining the relation between the diagonal coordinates of the vehicle and the lane lines according to the vehicle driving image information to be audited;

and when the diagonal coordinates of the vehicle are at the two ends of the lane line, the rechecking result shows that the vehicle is effective in line pressing running.

3. The method according to claim 1, wherein if the type of the illegal activity is reverse driving, performing the illegal activity judgment on the lane line information and the plurality of pieces of vehicle driving image information to be audited according to the illegal activity review algorithm to obtain a review result of the illegal activity, including:

determining a vehicle driving direction according to the plurality of pieces of vehicle driving image information to be checked;

and when the driving direction of the vehicle is opposite to the lane line direction in the lane line information, the review result is that the vehicle is effective to run reversely.

4. The method of claim 1, wherein if the type of violation is a stop violation or a road violation, performing a rule-breaking decision on the lane attribute and a plurality of pieces of vehicle driving image information to be checked according to the rule-breaking rule review algorithm, and determining a rule-breaking rule review result, including:

Determining lane attributes of the illegal behaviors according to the numbers of the image acquisition equipment;

when the lane attribute of the illegal act is inconsistent with the lane attribute in the vehicle illegal act information to be checked, the review result is invalid.

5. The method of claim 4, wherein if the lane attribute is a target vehicle type traffic lane, further comprising:

when the lane attribute of the illegal act is consistent with the lane attribute in the vehicle illegal act information to be audited, extracting license plate number information and vehicle type information in the vehicle driving image information;

when the license plate number information is consistent with the license plate number information of the illegal vehicle and the vehicle type information is consistent with the target vehicle type, the review result is that the illegal behavior is effective.

6. The method of claim 4, wherein if the lane attribute is a no-parking lane, further comprising:

when the lane attribute of the illegal act is consistent with the lane attribute in the vehicle illegal act information to be audited, extracting license plate number information in the vehicle driving image information;

when the license plate number information is consistent with the illegal vehicle license plate number information, judging whether vehicles corresponding to the license plate number information are in a forbidden parking area in a plurality of pieces of vehicle driving image information;

When the vehicles corresponding to the license plate number information are in the forbidden parking areas in the plurality of pieces of vehicle driving image information, the review result is that the illegal behaviors are effective.

7. The method of claim 4, wherein if the lane attribute is a single lane, further comprising:

when the license plate number information is consistent with the illegal vehicle license plate number information, positioning a target illegal vehicle;

determining the vehicle type of the target illegal vehicle according to the target model;

when the vehicle type is not the target vehicle type, the review result is that the illegal action is valid.

8. A vehicular illicit action review device, characterized by comprising:

the information acquisition module is used for acquiring the vehicle illegal action information to be audited, wherein the vehicle illegal action information comprises an illegal category, an image acquisition equipment number for acquiring the vehicle illegal action, a plurality of pieces of vehicle driving image information to be audited and illegal vehicle license plate number information, and the illegal category comprises: the method comprises the steps of driving in reverse, driving without a guide line, running red light, line pressing, stopping, distracted driving, unbuckling a safety belt and driving on a road in violation, wherein the image acquisition equipment number is used for acquiring lane line information or lane attributes of a pre-stored monitoring area corresponding to the image acquisition equipment number, and the lane attributes comprise a target vehicle type traffic limiting lane, a parking forbidden lane and a single-way lane;

The review algorithm determining module is used for determining a prestored illegal action review algorithm corresponding to the illegal category according to the illegal category;

the review result determining module is used for determining the review result of the illegal behaviors according to the illegal behavior review algorithm and the vehicle illegal behavior information; extracting license plate number information in the vehicle driving image information; when the extracted license plate number information is consistent with the illegal vehicle license plate number information, if the illegal type is retrograde, running without a guide line, line pressing running and illegal stopping, determining lane line information of illegal behaviors according to the number of the image acquisition equipment; carrying out illegal action judgment on the lane line information and the driving image information of a plurality of vehicles to be audited according to the illegal action review algorithm to obtain a review result of the illegal action; if the illegal variety is red light running, judging whether the traffic light is a red light currently according to the vehicle driving image information, if the traffic light is a red light currently, judging whether the vehicle position corresponding to the license plate number information exceeds a detection area, and if the vehicle position corresponding to the license plate number information exceeds the detection area, reviewing the result to be effective in illegal behaviors; if the illegal types are distracted driving or unbelted driving, vehicle detection is carried out according to a target vehicle positioning model to obtain a vehicle image corresponding to illegal vehicle license plate number information, personnel position detection is carried out on the vehicle image according to a target personnel positioning model to obtain a driver image, the driver image is input into a personnel behavior classification model to obtain a classification result, and a review result of illegal behaviors is determined according to the classification result.

9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the vehicle illicit review method of any of claims 1-7 when the program is executed by the processor.

10. A storage medium having stored thereon computer instructions which, when executed by a processor, implement the steps of the vehicle illicit behavior review method of any of claims 1-7.