CN113128540B

CN113128540B - Method and device for detecting vehicle theft behavior of non-motor vehicle and electronic equipment

Info

Publication number: CN113128540B
Application number: CN201911425301.8A
Authority: CN
Inventors: 蔡杭洲
Original assignee: Shenzhen Intellifusion Technologies Co Ltd
Current assignee: Shenzhen Intellifusion Technologies Co Ltd
Priority date: 2019-12-31
Filing date: 2019-12-31
Publication date: 2024-05-03
Anticipated expiration: 2039-12-31
Also published as: CN113128540A

Abstract

The embodiment of the invention provides a method and a device for detecting vehicle theft behavior of a non-motor vehicle and electronic equipment, wherein the method comprises the following steps: acquiring non-motor vehicle driving characteristics in a tracking image, wherein the non-motor vehicle driving characteristics comprise non-motor vehicle characteristics and driver characteristics matched with the non-motor vehicle characteristics; inputting the driving characteristics of the non-motor vehicle into a clustering engine, and clustering to obtain history snapshot information based on the driving characteristics of the non-motor vehicle, wherein the history snapshot information comprises corresponding history non-motor vehicle characteristics and history driver characteristics; detecting whether the driving characteristics of the non-motor vehicle are different from the corresponding historical snapshot information; and judging whether the driver has the vehicle stealing behavior according to the detection result. The vehicle theft behavior can be effectively detected in real time.

Description

Method and device for detecting vehicle theft behavior of non-motor vehicle and electronic equipment

Technical Field

The invention relates to the technical field of artificial intelligence, in particular to a method and a device for detecting vehicle theft behavior of a non-motor vehicle and electronic equipment.

Background

Image identification is one of the common technologies of current community security and traffic management, and can greatly improve social security sense, for example: and detecting traffic illegal events by using a community access control system based on face recognition or license plate image recognition. Because many non-motor vehicles do not deal with the license plate according to traffic management requirements, the attribution judgment of the non-motor vehicles is single, namely, who can open the lock of the non-motor vehicles, and who can drive the non-motor vehicles, the behavior of the non-motor vehicles is difficult to judge, so that the non-motor vehicles are easy to be stolen, and the vehicle-stealing personnel are difficult to find, and only after the vehicle owner alarms, the relevant personnel analyze a large amount of monitoring data, so that the corresponding clues can be found. Therefore, in the prior art, the vehicle theft behavior of the non-motor vehicle cannot be effectively detected in real time.

Disclosure of Invention

The embodiment of the invention provides a method and a device for detecting the vehicle theft behavior of a non-motor vehicle, electronic equipment and a computer readable storage medium, which can be used for effectively detecting the vehicle theft behavior of the non-motor vehicle in real time.

In a first aspect, an embodiment of the present invention provides a method for detecting vehicle theft behavior of a non-motor vehicle, including:

Acquiring non-motor vehicle driving characteristics in a tracking image, wherein the non-motor vehicle driving characteristics comprise non-motor vehicle characteristics and driver characteristics matched with the non-motor vehicle characteristics;

inputting the driving characteristics of the non-motor vehicle into a clustering engine, and clustering to obtain history snapshot information based on the driving characteristics of the non-motor vehicle, wherein the history snapshot information comprises corresponding history non-motor vehicle characteristics and history driver characteristics;

detecting whether the driving characteristics of the non-motor vehicle are different from the corresponding historical snapshot information or not, and obtaining a detection result;

And judging whether the vehicle theft behavior exists or not according to the detection result.

In a second aspect, an embodiment of the present invention provides a device for detecting vehicle theft behavior of a non-motor vehicle, including:

The first acquisition module is used for acquiring non-motor vehicle driving characteristics in the tracking image, wherein the non-motor vehicle driving characteristics comprise non-motor vehicle characteristics and driver characteristics matched with the non-motor vehicle characteristics;

The clustering module is used for inputting the driving characteristics of the non-motor vehicle into the clustering engine, and clustering to obtain history snapshot information based on the driving characteristics of the non-motor vehicle, wherein the history snapshot information comprises corresponding history non-motor vehicle characteristics and history driver characteristics;

The detection module is used for detecting whether the condition that the driving characteristics of the non-motor vehicle are different from the corresponding history snapshot information exists or not;

and the judging module is used for judging whether the vehicle theft behavior exists or not according to the detection result.

In a third aspect, an embodiment of the present invention provides an electronic device, including: the method comprises the steps of a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the steps in the method for detecting the vehicle theft behavior of the non-motor vehicle are realized when the processor executes the computer program.

In a fourth aspect, an embodiment of the present invention provides a computer readable storage medium, where a computer program is stored on the computer readable storage medium, where the computer program when executed by a processor implements steps in a method for detecting a vehicle theft behavior of a non-motor vehicle provided by the embodiment of the present invention.

In the embodiment of the invention, the driving characteristics of the non-motor vehicle in the tracking image are acquired, wherein the driving characteristics of the non-motor vehicle comprise the characteristics of the non-motor vehicle and the characteristics of a driver matched with the characteristics of the non-motor vehicle; inputting the driving characteristics of the non-motor vehicle into a clustering engine, and clustering to obtain history snapshot information based on the driving characteristics of the non-motor vehicle, wherein the history snapshot information comprises corresponding history non-motor vehicle characteristics and history driver characteristics; detecting whether the driving characteristics of the non-motor vehicle are different from the corresponding historical snapshot information; and judging whether the driver has the vehicle stealing behavior according to the detection result. Because the personnel driving the non-motor vehicles and the non-motor vehicles are identified and clustered, whether a situation of one person driving multiple vehicles or one vehicle driving multiple persons driving exists can be detected, whether the vehicle stealing behavior exists or not is judged according to the situations, analysis work of a large number of related personnel on monitoring data can be saved, and the vehicle stealing behavior can be effectively detected in real time.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic flow chart of a method for supervising traffic violations of non-motor vehicles according to an embodiment of the present invention;

FIG. 2 is a schematic flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention;

FIG. 3 is a schematic flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention;

FIG. 4 is a schematic flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention;

FIG. 5 is a schematic flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention;

FIG. 6 is a schematic flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention;

FIG. 7 is a schematic flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of a device for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention;

FIG. 9 is a schematic structural diagram of another device for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention;

FIG. 10 is a schematic structural diagram of another device for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention;

FIG. 11 is a schematic structural diagram of another device for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention;

fig. 12 is a schematic structural diagram of a first obtaining module according to an embodiment of the present invention;

fig. 13 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. 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.

Referring to fig. 1, fig. 1 is a flow chart of a method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention, as shown in fig. 1, including the following steps:

101. and acquiring the driving characteristics of the non-motor vehicle in the tracking image.

The non-motor vehicle driving feature comprises a non-motor vehicle feature and a driver feature matched with the non-motor vehicle feature, the non-motor vehicle driving feature can be understood as a feature that a driver drives the non-motor vehicle, and the tracking image refers to a continuous image containing the non-motor vehicle feature and the driver feature. The tracking image can be acquired by an image acquisition device, such as a community camera arranged in a community and at a community gate, or a traffic camera arranged at an intersection and a roadside, or a camera corresponding to the linkage of a community system and a traffic system, etc. The non-motor vehicle can be a bicycle, a tricycle, an electric vehicle, a balance vehicle, a manual vehicle and the like, and the non-motor vehicle features can be a vehicle head, a handle, a pedal, a tire and the like. The driver features may be facial features, such as facial contours, mouth, eyes, nose, etc. The matching of the non-motor vehicle characteristic and the driver characteristic can be performed according to the distance between the non-motor vehicle characteristic and the driver characteristic, or can be performed according to the non-motor vehicle characteristic and the movement track of the driver. The non-motor vehicle driving characteristics may be obtained by inputting the tracking image into a target tracking engine, extracting the tracking image by the target tracking engine, extracting the corresponding non-motor vehicle characteristics and driver characteristics of each image frame, and performing matching to obtain the matched non-motor vehicle characteristics and driver characteristics, i.e. obtaining the driving characteristics of the non-motor vehicle. Specifically, in the target tracking engine, a kalman filtering algorithm can be used for tracking the non-motor vehicle and the driver appearing in the video, and the corresponding non-motor vehicle characteristics and driver characteristics are extracted for matching, so that the non-motor vehicle driving characteristics are obtained.

It should be noted that the non-motor vehicle driving feature obtained in step 101 may include one or more non-motor vehicle features and one or more driver features. The tracking image may be referred to as a visual image, visual information, video image, video information, continuous video frame, continuous frame image, monitoring video, monitoring information, monitoring image, or the like.

In some possible embodiments, the above-mentioned driver features may further include a posture feature of the person, where the posture feature may be a hand posture, a foot posture, an upper body posture, or the like, and by analyzing the posture feature of the person, it may be determined whether the person is driving a non-motor vehicle, for example, whether the hand of the person is on the handle, whether the foot of the person is placed at the driving position, or the like, through the posture feature. The posture characteristics of the person can also be the posture characteristics of hands, feet, upper body, and the like of the non-motor vehicle matched with the driver and the passenger, for example, the driver and the passenger can be matched with the driver and the passenger of the non-motor vehicle by judging which person the hand on the handle belongs to, and the driver and the passenger of the non-motor vehicle can be matched with the posture characteristics of which person the foot on the pedal belongs to by judging the posture characteristics of which person, and of course, the posture characteristics are required to be matched with the driver and the passenger of the non-motor vehicle when the driver and the passenger of the non-motor vehicle are matched with the posture characteristics, so that the driver and the passenger can be determined. The gesture features may be associated with the face features to match the gesture features to the corresponding person.

102. And inputting the driving characteristics of the non-motor vehicle into a clustering engine, and clustering to obtain historical snapshot information based on the driving characteristics of the non-motor vehicle.

In this step, the non-motor vehicle driving characteristics include non-motor vehicle characteristics and driver characteristics matched with the non-motor vehicle characteristics, and the history snapshot information includes corresponding history non-motor vehicle characteristics and history driver characteristics. The above-mentioned history snapshot information may be history snapshot information within a preset time, such as snapshot information within 1 month, or snapshot information within 6 months. The non-motor vehicle driving feature may be one or more, i.e. the non-motor vehicle feature may be one or more, the driver feature may be one or more, and the non-motor vehicle feature matches only one driver feature. Such as: the tracking image comprises non-motor vehicle features b1, b2, b3 and b4, and personnel features a1, a2, a3, a4, a5 and a6, wherein the non-motor vehicle driving features are 1,2, 3 and 4, the non-motor vehicle driving feature 1 can be represented as matching of the non-motor vehicle feature b1 and the personnel feature a1, at the moment, the personnel with the personnel feature a1 can be called a driver, and the personnel feature a1 is the driver feature a1; the non-motor vehicle driving feature 2 may be represented as a match of the non-motor vehicle feature b2 with the person feature a2, at which time the person having the person feature a2 may be referred to as a driver, and the person feature a2 is the driver feature a2; by analogy, the person feature a3 and the person feature a4 correspond to the driver features a3 and a4, respectively; a5 and a6 are not driver characteristics. The clustering engine can cluster the historical snapshot information of the driver according to the characteristic of the driver, and can cluster the snapshot information of the non-motor vehicle according to the characteristic of the non-motor vehicle, wherein the clustering can be understood as gathering objects with the same characteristic. The above-mentioned history snapshot information may be the history snapshot information that the collecting device captures to the target non-motor vehicle feature or the target driver feature in step 101, and the history snapshot information may be the fixed-length or variable-length semi-structured data stored in the history database. In the clustering process, semi-structured data related to the target features are extracted from the historical database according to the corresponding target features to perform clustering. For example, the driver characteristics corresponding to the drivers A1 and A2 are A1 and A2, the non-motor vehicle characteristics corresponding to the non-motor vehicles B1 and B2 driven by the drivers A1 and A2 are B1 and B2, and the history snapshot information of the related driver A1 can be extracted from the history database according to the driver characteristic A1 for clustering; the history snapshot information of the related driver A2 can be extracted from the history database according to the driver characteristics A2 for clustering; the history snapshot information of the related non-motor vehicle B1 can be extracted from the history database according to the non-motor vehicle characteristics B1 for clustering; the historical snapshot information of the related non-motor vehicle B2 can be extracted from the historical database according to the non-motor vehicle characteristics B2 for clustering.

It should be noted that the above-mentioned historical non-motor vehicle features may be non-motor vehicle features such as a vehicle head, a handle, a pedal, a tire, and the like. The historical driver characteristics may be historical face characteristics, such as facial contours, mouth, eyes, nose, and the like.

103. And detecting whether the driving characteristics of the non-motor vehicle are different from the corresponding historical snapshot information, and obtaining a detection result.

In this step, the non-motor vehicle driving characteristics include non-motor vehicle characteristics and driver characteristics matched with the non-motor vehicle characteristics, and the history snapshot information includes corresponding history non-motor vehicle characteristics and history driver characteristics. The above-mentioned detection may be to detect whether the corresponding driver feature is the same as the historical driver feature under the same non-motor vehicle feature, or to detect whether the corresponding non-motor vehicle feature is the same as the historical non-motor vehicle feature under the same driver feature. For example, assuming that the above-mentioned non-motor vehicle driving characteristic is (A1, B1), the above-mentioned A1 corresponds to the driver characteristic of the driver A1, the above-mentioned B1 corresponds to the non-motor vehicle characteristic of the non-motor vehicle B1, and the above-mentioned driver A1 drives the non-motor vehicle B1, if the same non-motor vehicle characteristic is detected, whether the corresponding driver characteristic is the same as the historical driver characteristic may be: if the history snapshot information corresponding to B1 is detected to be (A1, B1), the driver corresponding to the non-motor vehicle characteristic B1 representing the non-motor vehicle B1 is always A1, that is, the corresponding driver characteristic is the same as the history driver characteristic. If the history snapshot information corresponding to b1 is detected to be (a 2, b 1), (a 3, b 1), (a 4, b 1) and (a 4, b 1), the driver corresponding to the non-motor vehicle feature b1 is not always A1, that is, the corresponding driver feature is different from the history driver feature. Under the condition that the same driver characteristic is detected, whether the corresponding non-motor vehicle characteristic is equal to the historical non-motor vehicle characteristic or not can be: if the history snapshot information corresponding to the A1 is detected to be (A1, B1), (A1, B1) and (A1, B1), the non-motor vehicle corresponding to the driver characteristic A1 of the driver A1 is always B1, that is, the non-motor vehicle corresponding to the driver A1 is the same as the history non-motor vehicle characteristic. If the history snapshot information corresponding to the A1 is detected to be (A1, b 2), (A1, b 3), (A1, b 4) and (A1, b 5), the non-motor vehicle corresponding to the driver characteristic A1 of the driver A1 is always changed, namely the non-motor vehicle corresponding to the driver A1 is different from the history non-motor vehicle characteristic. The above-mentioned detection may be that the non-motor vehicle feature and the historical non-motor vehicle feature are compared one by one in similarity, when the similarity of the historical non-motor vehicle feature and the non-motor vehicle feature is greater than or equal to a non-motor vehicle similarity threshold, the non-motor vehicle feature may be considered to be the same as the historical non-motor vehicle feature, and when the similarity of the historical non-motor vehicle feature and the non-motor vehicle feature is less than the non-motor vehicle similarity threshold, the non-motor vehicle feature may be considered to be different from the historical non-motor vehicle feature. Similarly, the above-mentioned detection may be to compare the driver feature with the historical driver feature one by one, where when the similarity between the historical driver feature and the driver feature is greater than or equal to the driver similarity threshold, the driver feature may be considered to be the same as the historical driver feature, and when the similarity between the historical driver feature and the driver feature is less than the driver similarity threshold, the driver feature may be considered to be different from the historical driver feature.

104. And judging whether the vehicle theft behavior exists or not according to the detection result.

The above detection results include four kinds of steps 103, the first kind: under the same non-motor vehicle characteristics, the corresponding driver characteristics are the same as the historical driver characteristics; second kind: under the same non-motor vehicle characteristics, the corresponding driver characteristics are different from the historical driver characteristics; third kind: under the same driver characteristics, the corresponding non-motor vehicle characteristics are the same as the historical non-motor vehicle characteristics; fourth kind: under the same driver characteristics, the corresponding non-motor vehicle characteristics are different from the historical non-motor vehicle characteristics. In the first detection result, since the driver characteristic is the same as the history driver characteristic, it can be explained that the vehicle corresponding to the non-vehicle characteristic is always driven by the driver corresponding to the driver characteristic, and it can be considered that there is no theft behavior. In the second detection result, since the driver characteristics are different from the historical driver characteristics, it can be explained that the motor vehicle corresponding to the non-motor vehicle characteristics is always driven by different drivers, and the non-motor vehicle is always turned by hands, so that the stolen vehicle behavior can be considered. In the third detection result, since the non-motor vehicle feature is the same as the history non-motor vehicle feature, it can be explained that the driver corresponding to the driver feature always drives the same non-motor vehicle, and it can be considered that no theft behavior exists. In the fourth detection result, since the non-motor vehicle characteristic is different from the history non-motor vehicle characteristic, it can be explained that the driver corresponding to the driver characteristic always drives different non-motor vehicles, and it can be considered that the vehicle theft behavior exists.

In some possible embodiments, when the third detection result is obtained, further identifying the non-motor vehicle, and identifying whether the non-motor vehicle is a shared non-motor vehicle, if the non-motor vehicle is identified as the non-shared non-motor vehicle, determining that there is a vehicle theft behavior, and if the non-motor vehicle is identified as the shared non-motor vehicle, determining that there is no vehicle theft behavior.

It should be noted that, the method for detecting the vehicle theft behavior of the non-motor vehicle provided by the embodiment of the invention can be applied to the device for detecting the vehicle theft behavior of the non-motor vehicle, for example: and equipment such as community cameras, traffic cameras, computers and servers for detecting the vehicle theft behavior of the non-motor vehicles.

In the embodiment of the invention, as the personnel driving the non-motor vehicles and the non-motor vehicles are identified and clustered, whether one person drives multiple vehicles or one vehicle drives multiple persons can be detected, so that whether the vehicle theft behavior exists or not can be judged according to the conditions, analysis work of a large number of related personnel on monitoring data can be saved, and the vehicle theft behavior can be effectively detected in real time.

Optionally, referring to fig. 2, fig. 2 is a flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention, as shown in fig. 2, including the following steps:

201. and acquiring the driving characteristics of the non-motor vehicle in the tracking image.

Wherein the non-motor vehicle driving characteristics include non-motor vehicle characteristics and driver characteristics matched with the non-motor vehicle characteristics.

202. And inputting the driver characteristics into a clustering engine, and clustering to obtain historical snapshot information based on the driver characteristics.

The historical snapshot information comprises corresponding historical non-motor vehicle characteristics and historical driver characteristics.

203. And detecting whether the non-motor vehicle characteristics corresponding to the same driver characteristics are different from the historical non-motor vehicle characteristics.

204. If the non-motor vehicle features corresponding to the same driver features are detected to be different from the historical non-motor vehicle features, judging that the driver has the vehicle stealing behavior.

In this embodiment, step 202 clusters driver features, which may be face features, and the face features may also be referred to as face images, and the face features are input into a clustering engine, and history snapshot data with the same face features are extracted from a history database by the clustering engine. Because the clustering is completed based on the characteristics of the drivers, each class is distinguished by the characteristics of the drivers, and each class is understood to be the historical snapshot information corresponding to the same driver. For example, the driver features are A1, A2, and A3, respectively corresponding to the drivers A1, A2, and A3, the history snapshot information of the driver A1 is obtained according to the clustering of A1, the history snapshot information of the driver A2 is obtained according to the clustering of A2, and the history snapshot information of the driver A3 is obtained according to the clustering of A3.

In the detection in step 203, the non-motor vehicle features corresponding to the same driver may be compared with the historical non-motor vehicle features in the historical snapshot information one by one, so as to determine whether the non-motor vehicle features are the same as the historical non-motor vehicle features. Specifically, when there is a similarity of a historical non-motor vehicle feature to the non-motor vehicle feature that reaches a similarity threshold (e.g., 99%), the historical non-motor vehicle feature may be considered the same as the non-motor vehicle feature; when there is a similarity of a historical non-motor vehicle feature to the non-motor vehicle feature that does not reach a similarity threshold (e.g., 99%), the historical non-motor vehicle feature may be considered to be different from the non-motor vehicle feature. And detecting each driver in the tracking image, so as to obtain a detection result of each driver.

In step 204, since the corresponding non-motor vehicle feature is different from the historical non-motor vehicle feature under the same driver feature, it may be stated that the driver corresponding to the driver feature always drives different non-motor vehicles, and it may be considered that the driver has a vehicle theft behavior.

In the step, the historical snapshot information of the same driver is obtained through the clustering of the characteristics of the driver, and whether the driver has the vehicle theft behavior can be analyzed. In addition, when the same driver is driving different non-motor vehicles, the driver can be considered to have vehicle theft behavior. Through clustering the driver characteristics, whether the driver has the vehicle theft behavior can be detected rapidly, and the detection efficiency is improved.

Optionally, referring to fig. 3, fig. 3 is a flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention, as shown in fig. 3, including the following steps:

301. And acquiring the driving characteristics of the non-motor vehicle in the tracking image.

302. And inputting the non-motor vehicle characteristics into a clustering engine, and clustering to obtain historical snapshot information based on the non-motor vehicle characteristics.

303. And detecting whether the situation that the driver characteristics corresponding to the same non-motor vehicle are different from the historical driver characteristics exists.

304. If the fact that the characteristics of the drivers corresponding to the same non-motor vehicle are different from the characteristics of the historical drivers is detected, the vehicle theft behavior is judged.

In this embodiment, step 302 clusters the non-motor vehicle features, inputs the non-motor vehicle features into the clustering engine, extracts the history snapshot data with the same non-motor vehicle features from the history database through the clustering engine, and when a plurality of non-motor vehicles corresponding to driving exist for the driver, can also establish different non-motor vehicle files according to different non-motor vehicle features for storing the extracted corresponding history snapshot data, thereby completing the clustering based on the non-motor vehicle features. Because the clustering is completed based on the characteristics of the non-motor vehicles, each category is distinguished by the characteristics of the non-motor vehicles, and each category is understood to be the historical snapshot information corresponding to the same non-motor vehicle. For example, the characteristics of the non-motor vehicles are B1, B2 and B3 respectively corresponding to the non-motor vehicles B1, B2 and B3, the historical snapshot information of the non-motor vehicle A1 is obtained by clustering according to the B1, the historical snapshot information of the non-motor vehicle B2 is obtained by clustering according to the B2, and the historical snapshot information of the non-motor vehicle B3 is obtained by clustering according to the B3.

In the above detection in step 303, the driver features corresponding to the same non-motor vehicle may be compared with the historical driver features in the historical snapshot information one by one, so as to determine whether the driver features are the same as the historical driver features. Specifically, when the similarity between the historical driver feature and the driver feature reaches a similarity threshold (e.g., 99%), the historical driver feature may be considered to be the same as the driver feature; when there is a similarity of a historical driver characteristic to the driver characteristic that does not reach a similarity threshold (e.g., 99%), the historical driver characteristic may be considered to be different from the driver characteristic. And detecting each non-motor vehicle in the tracking image, so as to obtain a detection result of each non-motor vehicle. The driver characteristic may be a face characteristic, and the history driver characteristic may be a face characteristic captured by a history.

In step 304, since the corresponding driver characteristic is different from the historical driver characteristic under the same non-motor vehicle characteristic, it can be stated that the non-motor vehicle corresponding to the non-motor vehicle characteristic is always driven by different drivers, and it can be considered that the non-motor vehicle has a theft condition (frequent turning of the non-motor vehicle), and at the same time, it can be considered that the driver driving the stolen non-motor vehicle has a theft behavior, and the historical driver also has a theft behavior.

In some possible embodiments, since the shared non-motor vehicles also have different driving situations of drivers, the non-motor vehicles can be identified before judging whether the vehicle-stealing behavior exists, whether the non-motor vehicles are shared non-motor vehicles can be judged, if the non-motor vehicles are shared non-motor vehicles, the vehicle-stealing behavior judgment is not made, or the vehicle-stealing behavior is directly considered to be absent, so that the computing resource is saved. In other possible embodiments, the identification and judgment of the shared non-motor vehicles can be judged in the tracking image, so that clustering of the shared non-motor vehicles can be avoided, and the computing resources are further saved.

In the step, the non-motor vehicle features are clustered to obtain the history snapshot information of the same non-motor vehicle, and whether the non-motor vehicle is stolen or not can be analyzed. In addition, when the same non-motor vehicle is driven by different drivers, the non-motor vehicle can be considered to be stolen. By clustering the characteristics of the non-motor vehicles, whether the non-motor vehicles are stolen or not can be detected rapidly, so that vehicle theft personnel are analyzed, and the detection efficiency is improved.

Optionally, referring to fig. 4, fig. 4 is a flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention, as shown in fig. 4, including the following steps:

401. And acquiring the driving characteristics of the non-motor vehicle in the tracking image.

The driving characteristics of the non-motor vehicle comprise the characteristics of the non-motor vehicle and the characteristics of a driver matched with the characteristics of the non-motor vehicle, and the characteristics of the driver comprise the characteristics of a face.

402. And inputting the driving characteristics of the non-motor vehicle into a clustering engine, and clustering to obtain historical snapshot information based on the driving characteristics of the non-motor vehicle.

403. And detecting whether the driving characteristics of the non-motor vehicle are different from the corresponding historical snapshot information.

404. And judging whether the vehicle theft behavior exists or not according to the detection result.

405. And after judging that the vehicle theft behavior exists, acquiring corresponding peer information of drivers with the vehicle theft behavior.

406. And constructing a relation map of the drivers with the vehicle theft behaviors according to the peer information.

The peer information in step 405 refers to the person information in the same image as the driver is snapped, and the peer information may be walking or driving a non-motor vehicle peer. The peer information may be peer information of a preset time, for example, peer information within 1 month, peer information within 6 months, and the like. The peer information includes peer times and peer personnel information. The peer information may be obtained in a peer database, where the peer database stores peer files of the person to be captured, and peer image data of the person to be captured and the peer person are recorded in the peer files, where the peer image data includes peer locations, peer times, and the like. Specifically, the face features of the drivers with the vehicle theft behaviors are obtained, and the peer files of the drivers are searched in the peer database through the face features of the drivers, so that corresponding peer information is obtained.

In step 406, the relationship graph includes the number of fellow passenger and the information of the fellow passenger, where the relationship graph is used to show the fellow passenger's relationship with the driver who has the vehicle theft behavior. The relationship map may be a connection relationship image formed by taking a face image of a driver as a center point and a face image of a peer person as a connection point, each connection line represents a peer relationship between the driver and a peer person, and the peer times may be displayed on the connection line or on a face image edge of the corresponding peer person.

In one possible embodiment, in the relationship map, only the fellow passenger with the driver's fellow passenger number exceeding the preset number threshold is displayed, for example, only the fellow passenger with the driver's fellow passenger number exceeding 5 times is displayed.

It should be noted that, the above steps 405 and 406 are optional, and in some possible embodiments, only the vehicle theft behavior needs to be determined as a basis for the determination. The above steps 405 and 406 may be combined with the embodiments corresponding to fig. 2 and 3.

In the steps, the relationship map of the vehicle-stealing personnel is formed by acquiring the peer information of the vehicle-stealing personnel (namely the drivers with vehicle-stealing behaviors), and the social relationship of the vehicle-stealing personnel can be intuitively displayed, so that the analysis and excavation of the vehicle-stealing personnel group by the relevant police personnel are facilitated.

Optionally, referring to fig. 5, fig. 5 is a flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention, as shown in fig. 5, including the following steps:

501. And acquiring the driving characteristics of the non-motor vehicle in the tracking image.

502. And inputting the driving characteristics of the non-motor vehicle into a clustering engine, and clustering to obtain historical snapshot information based on the driving characteristics of the non-motor vehicle.

503. And detecting whether the driving characteristics of the non-motor vehicle are different from the corresponding historical snapshot information.

504. And judging whether the vehicle theft behavior exists or not according to the detection result.

505. And after judging that the vehicle theft behavior exists, acquiring corresponding peer information of drivers with the vehicle theft behavior.

506. And constructing a relation map of the drivers with the vehicle theft behaviors according to the peer information.

507. And comparing the staff information with the staff library of the front department.

508. And (3) adjusting the relation map according to the comparison result to obtain a suspected person relation map with vehicle theft behavior.

In step 507, the front department personnel library may include front department information of vehicle theft personnel, further, the front department information of vehicle theft personnel refers to front department information of vehicle theft personnel for stealing non-motor vehicles, the front department information includes vehicle theft personnel information and vehicle theft behavior information, the vehicle theft behavior information includes vehicle theft time, vehicle theft place, vehicle type and the like, and the vehicle theft personnel information includes face feature, identity information and the like. The identity information can be information such as name, identification card number, age, birth date and the like. The peer information includes face characteristics, identity information and the like. The comparison may be of face features or identity information. Specifically, the face features of the fellow staff can be compared with the face features of the vehicle-stealing staff in the front department staff library, or the identity information of the fellow staff can be compared with the identity information of the vehicle-stealing staff in the front department staff library.

In step 507, after the peer person information is compared in the front department personnel library, if the peer person corresponding to the peer person information is a front department person in the front department personnel library, the relationship map is adjusted, where the adjustment may be to highlight the connection line between the peer person and the driver, for example, the connection line with a gray common peer relationship is a red connection line with the front department person, or directly mark the face image of the peer person, and mark the face image as the front department person, thereby obtaining the relationship map of suspected person. The suspicion relationship map is used for displaying the staff with larger suspicion of vehicle theft, such as front department staff, for example, staff with more than 50 times of staff.

It should be noted that, the steps 505 and 506 are optional, and in some possible embodiments, only the vehicle theft behavior needs to be determined as a basis for the determination. The above-described steps 505 and 506 may be combined with the embodiments corresponding to fig. 2 and 3.

In the steps, the relationship map of the vehicle-stealing personnel is formed by acquiring the peer information of the vehicle-stealing personnel (namely the drivers with vehicle-stealing behaviors), and the social relationship of the vehicle-stealing personnel can be intuitively displayed, so that the analysis and excavation of the vehicle-stealing personnel group by the relevant police personnel are facilitated. In addition, whether the staff is a front department staff is inquired, a suspected staff relationship map is obtained, and the social relationship of the vehicle theft staff is further intuitively displayed, so that the vehicle theft staff is convenient to further analyze and excavate.

Optionally, referring to fig. 6, fig. 6 is a flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention, as shown in fig. 6, including the following steps:

601. And acquiring the driving characteristics of the non-motor vehicle in the tracking image.

602. And inputting the driving characteristics of the non-motor vehicle into a clustering engine, and clustering to obtain historical snapshot information based on the driving characteristics of the non-motor vehicle.

603. And detecting whether the driving characteristics of the non-motor vehicle are different from the corresponding historical snapshot information.

604. And judging whether the vehicle theft behavior exists or not according to the detection result.

605. After judging that the vehicle theft behavior exists, continuously taking a snapshot of a driver with the vehicle theft behavior, and acquiring corresponding snapshot information.

606. And forming a track of the driver aiming at the vehicle theft behavior according to the snapshot information.

In step 605, continuously taking a snapshot of the driver who has the stolen vehicle behavior can be understood as tracking the driver, that is, taking a snapshot of the driver by cameras arranged in a plurality of places. The snapshot information comprises the face characteristics, the snapshot time and the snapshot place of the driver, and the snapshot place can be equal to the installation place of the camera.

In step 606, a track of the driver can be obtained according to the capturing time and the capturing place, the track is used for analyzing the place where the driver appears, and the track can be understood as what the moving route is in what time period the driver moves from what place to what place. For example, the snapshot time is 12 points and 11 minutes, and the snapshot place is the back door of a certain district; the snapshot time is 12 points and 30 minutes, and the snapshot place is an intersection A; the snapshot time is 12 points and 50 minutes, and the snapshot place is the intersection B; the snapshot time is 13 points and 15 minutes, and the snapshot place is a certain handcart line. The resulting trajectory is: rear door of certain district- & gtA road junction- & gtB road junction- & gtcertain second-hand vehicle. The vehicle theft personnel can be analyzed through the tracks, such as analyzing where the dirty points are located, analyzing where the gathering points of the group personnel are located, and the like. For example, at the time of the intersection a, the driver also drives the stolen non-motor vehicle, and at the time of the intersection B, the driver does not drive the stolen non-motor vehicle, so that the dirty point can be obtained between the intersection a and the intersection B, and the more accurate position of the dirty point can be calculated by calculating the speed of the driver at the intersection a and the speed of the driver at the intersection B and combining the time from the intersection a to the intersection B and the distance from the intersection a to the intersection B. The position of the gathering point of the partner can be judged to be near a certain place by the fact that the driver drives the stolen non-motor vehicle to be in the same way as other people and disappears after the same way arrives at the certain place.

It should be noted that, the above-mentioned step 605 and step 606 are optional, and in some possible embodiments, only the vehicle theft behavior needs to be determined as the basis for the determination. The above-described step 605 and step 606 may be combined with the embodiments corresponding to fig. 2 and 3.

In the steps, continuous snapshot is carried out on the vehicle theft personnel (refer to the drivers with vehicle theft behaviors) to obtain the track of the vehicle theft personnel, and the moving track of the vehicle theft personnel can be intuitively displayed, so that analysis and excavation of the gathering points or the dirty selling points of the vehicle theft personnel are conveniently carried out by the relevant police personnel.

Optionally, referring to fig. 7, fig. 7 is a flow chart of another method for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention, as shown in fig. 7, including the following steps:

701. and acquiring non-motor vehicle motion information and personnel motion information in the tracking image.

702. And matching the non-motor vehicle characteristics with corresponding driver characteristics according to the non-motor vehicle movement information and the personnel movement information to form non-motor vehicle driving characteristics.

703. And inputting the driving characteristics of the non-motor vehicle into a clustering engine, and clustering to obtain historical snapshot information based on the driving characteristics of the non-motor vehicle.

704. And detecting whether the driving characteristics of the non-motor vehicle are different from the corresponding historical snapshot information.

705. And judging whether the vehicle theft behavior exists or not according to the detection result.

The motion information of the non-motor vehicle feature may be a speed or a track of the non-motor vehicle feature, and the motion information of the face feature may be a speed or a track of the face feature. In the image recognition technology, the feature to be recognized is made into a feature frame, for example, a non-motor vehicle feature frame and a face feature frame are made respectively, the motion information of the non-motor vehicle feature frame can be embodied as the motion information of the non-motor vehicle feature frame, the motion information of the face feature frame can be embodied as the motion information of the face feature frame, the face feature in the non-motor vehicle feature preset range can be obtained by obtaining the face feature frame in the non-motor vehicle feature preset range, the preset range can be the pixel radius of the non-motor vehicle feature frame on the image, the size radius can also be the size radius, for example, the face feature in the non-motor vehicle feature 100-pixel radius can be obtained, and for example, the face feature frame with the distance from the center of the face feature frame to the center of the non-motor vehicle feature frame smaller than 200 pixels can be obtained. The face features closest to or identical to the motion information of the non-motor vehicle features can be extracted through the motion information of the non-motor vehicle features and the motion information of the face features within the preset range of the non-motor vehicle features and can be used as the face features owned by the driver, so that the driver can be matched. For example, assuming that the non-motor vehicle features are B and the motion information of B is C, the motion information of faces A1, A2, A3, A1, A2, A3 are D1, D2, D3 respectively in the preset range of B, one of D1, D2, D3 closest to C is selected, if D1 is closest to C, the face feature corresponding to D1 is A1, the person matching the face feature A1 is a driver, and C, D, D2, D3 in the example may be speeds, tracks, or distances.

Optionally, the motion information according to the feature of the non-motor vehicle and the motion information of the face feature in the preset range is a non-locomotive matching driver, including: forming a face set according to the motion information of the non-motor vehicle features and the motion information of the face features in a preset range; and matching the non-locomotive with the driver according to the face set.

Specifically, the motion information of the non-motor vehicle features may be the speed or the track of the non-motor vehicle features, the motion information of the face features may be the speed or the track of the face features, the face set is a face feature set within a preset range of the non-motor vehicle features, the face set may be identified by a specific number or letter label, for example, a face set E, a face set 1, etc., or may be identified by a feature value of the non-motor vehicle features, for example, a non-motor vehicle feature is B, the face set may be identified as a face set B, or may be identified by a color of a feature frame on an image, for example, a feature frame of a non-motor vehicle feature is green on the image, a green frame is also made for the feature frame of the face features within the preset range of the non-motor vehicle feature, and the face set may be identified by a green representative identifier G. In step 202, visual tracking is performed on the non-motor vehicle features and the face features, and a multi-frame image may be obtained. The motion information of the non-motor vehicle features and the motion information of the face features can be calculated through the multi-frame images, and the face features, close to the motion information of the non-motor vehicle features, of the motion information are extracted within a preset range of the non-motor vehicle features, so that a face set is formed. The preset range can be the radius of a pixel on an image, the radius of a dimension, the motion information of the face feature is similar to the motion information of the non-motor vehicle feature, the motion information threshold of the non-motor vehicle feature can be used, for example, the speed threshold interval of the non-motor vehicle feature can be used, and when the speed of the face feature falls into the speed threshold interval of the non-motor vehicle feature, the motion information of the face feature can be regarded as similar to the motion information of the non-motor vehicle feature. For example: the speed threshold interval of the non-motor vehicle features is 30 codes to 35 codes, and when the speed of the face features is 33 codes, the speed of the face features and the speed of the non-motor vehicle features can be regarded as similar, so that the motion information of the face features and the motion information of the non-motor vehicle features are regarded as similar. The speed threshold interval of the non-motor vehicle feature may be set by the speed of the non-motor vehicle feature. The motion information of the face features is similar to the motion information of the non-motor vehicle features, or the motion information of the non-motor vehicle features and the track similarity threshold of the face features can be also used, when the track similarity of the face features and the track similarity of the non-motor vehicle features reach the track similarity threshold, the motion information of the face features and the motion information of the non-motor vehicle features can be regarded as similar. For example: the track similarity threshold of the non-motor vehicle features and the human face features can be set to 80%, and when the track similarity of the human face features and the track similarity of the non-motor vehicle features reaches more than 80%, the motion information of the human face features and the motion information of the non-motor vehicle features can be regarded as similar.

In some possible embodiments, a face set may be formed in each frame of image, multiple frames of images may obtain multiple face sets, and intersections of the multiple face sets may be obtained to obtain a final face set. The face features in the final face set obtained by intersection are face features which always exist in a preset range of the features of the non-motor vehicle in the visual tracking process, and can also be said to be face features which keep a certain distance when the non-motor vehicle runs through Cheng Yufei motor vehicles. For example: assuming that 400 frames of images are obtained in visual tracking, 4 frames of images including the 100 th frame, the 200 th frame, the 300 th frame and the 400 th frame are extracted to form 4 face sets E1, E2, E3 and E4 respectively, the face features A1, A2, A3 and A4 are included in the E1, the face features A1, A2, A4 and A5 are included in the E2, the face features A1, A2, A5 and A6 are included in the E4, the face features A1 and A2 are included in the final face set E of the intersection set of the E1, the E2, the E3 and the E4, and a driver can be matched in the final face set E, and at the moment, the face features closest to the motion information of the non-motor vehicle features can be matched as the face features possessed by the driver by comparing the motion information of the face features A1 and A2 with the motion information of the non-motor vehicle features.

Optionally, the motion information of the non-motor vehicle feature includes a speed of the non-motor vehicle feature, and the motion information of the face feature includes a speed of the face feature; forming a face set according to the motion information of the non-motor vehicle features and the motion information of the face features in a preset range, including: comparing the speed of the non-motor vehicle characteristic with the speed of the face characteristic to obtain a speed comparison result; and forming the face set according to the speed comparison result.

Specifically, visual tracking is performed on the non-motor vehicle features and the face features, so that multi-frame images can be obtained. The speed of the non-motor vehicle characteristic and the speed of the face characteristic can be calculated through the multi-frame images, and the face characteristic with the speed similar to the speed of the non-motor vehicle characteristic is extracted within the preset range of the non-motor vehicle characteristic to form a face set. The speed of the non-motor vehicle feature and the speed of the face feature can be obtained by calculating the position change and time of the center point of the feature frame in the image by the image recognition technology, the preset range can be the pixel radius on the image, the size radius can also be the size radius, the speed of the face feature is similar to the speed of the non-motor vehicle feature, the speed threshold of the non-motor vehicle feature can be used for example, the speed threshold of the non-motor vehicle feature can be used for comparing the speed of the face feature with the speed threshold of the non-motor vehicle feature, and when the speed of the face feature falls into the speed threshold of the non-motor vehicle feature, the speed of the face feature can be regarded as similar to the speed of the non-motor vehicle feature, for example: the speed threshold interval of the non-motor vehicle features is 30 codes to 35 codes, the speed of the face feature A1 is 33 codes, the speed of the face feature A2 is 34 codes, the speed of the face feature A3 is 29 codes, the speeds of the face features A1 and A2 can be regarded as similar to the speed of the non-motor vehicle features, and accordingly the face features A1 and A2 are recorded and enter the face set, people which are obviously not drivers of the motor vehicle, such as people with the face feature A3, only the face features A1 and A2 in the face set are removed, face feature elements in the face set are reduced, only the face features A1 and A2 in the face set need to be calculated during matching, and the calculated amount of matching the drivers through the face set is reduced. When the formed face set is an empty set, the face can be integrated into the non-empty face set by increasing the preset range of the non-motor vehicle features and/or expanding the speed threshold interval of the non-motor vehicle features.

Optionally, the motion information of the non-motor vehicle feature includes a track of the non-motor vehicle feature, and the motion information of the face feature includes a track of the face feature; forming a face set according to the motion information of the non-motor vehicle features and the motion information of the face features in a preset range, including: comparing the track of the non-motor vehicle characteristic with the track of the face characteristic to obtain a track comparison result; and forming the face set according to the track comparison result.

Specifically, the track comparison result may be the track coincidence ratio of the track of the face feature and the track of the non-motor vehicle feature, for example, the track of the center of the feature frame of the face feature in the continuous image is compared with the track of the center of the feature frame of the non-motor vehicle feature in the continuous image with the same end point, and the ratio of the track coincidence length of the face feature and the track of the non-motor vehicle feature to the track total length of the track of the face feature and the track of the non-motor vehicle feature is calculated to obtain the coincidence ratio, for example: assuming that the track length of the face feature is 49, the track length of the non-motor vehicle feature is 51, the overlap length is 45, and the overlap ratio is 45×2/(49+51) ×100% =90%. If the contact ratio 90% is greater than the contact ratio threshold of the tracks of the face features and the tracks of the non-motor vehicle features, the face features can be recorded into the face set.

In addition, the track comparison result may also be a comparison result of track equations of the track of the face feature and the track of the non-motor vehicle feature, for example, a constant ratio or a difference value in two track equations may be used, the closer the ratio is to 1, the more similar the two track equations are, the closer the difference value is to 0, and the more similar the two track equations are.

Optionally, the matching the driver for the non-locomotive according to the face set includes: detecting the number of the face features of the face set, and judging whether the face set has a plurality of face features or not; and if a plurality of face features exist, selecting the face feature closest to the non-motor vehicle feature pixels from the plurality of face features, and determining the face feature as the face feature of the driver to obtain the driver of the non-motor vehicle.

Specifically, the pixel may be the smallest pixel interval between the center of the feature frame of the face feature and the center of the feature frame of the non-motor vehicle feature in the image, or the smallest pixel interval between the edge of the feature frame of the face feature and the center of the feature frame of the non-motor vehicle feature.

In addition, if there are no more face features in the face set, that is, the face features in the face set are unique, the unique face features in the face set can be directly considered to belong to the driver, so as to determine the driver.

In some possible embodiments, if a plurality of face features exist in the face set, the face feature closest to the feature size of the non-motor vehicle may be selected from the plurality of face features to be determined as the face feature of the driver, so as to obtain the driver of the non-motor vehicle. In addition, the face feature with the highest image quality can be selected to be determined as the face feature of the driver.

In the steps, the non-motor vehicle is matched with the driver, so that the non-motor vehicle is matched with the only driver, and the false detection caused by the participation of other people can be avoided.

Referring to fig. 8, fig. 8 is a schematic structural diagram of a device for detecting vehicle theft behavior of a non-motor vehicle according to an embodiment of the present invention, where, as shown in fig. 8, the device includes:

A first obtaining module 801, configured to obtain a non-motor vehicle driving feature in a tracking image, where the non-motor vehicle driving feature includes a non-motor vehicle feature and a driver feature matched with the non-motor vehicle feature;

The clustering module 802 is configured to input the non-motor vehicle driving feature into a clustering engine, and cluster to obtain history snapshot information based on the non-motor vehicle driving feature, where the history snapshot information includes a corresponding history non-motor vehicle feature and a history driver feature;

The detection module 803 is configured to detect whether there is a situation in which the driving characteristics of the non-motor vehicle are different from the corresponding history snapshot information;

and the judging module 804 is configured to judge whether there is a vehicle theft behavior according to the detection result.

Optionally, as shown in fig. 8, the clustering module 802 is further configured to input the driver features into a clustering engine, and cluster to obtain history snapshot information based on the driver features;

The detection module 803 is further configured to detect whether there is a situation in which a non-motor vehicle feature corresponding to the same driver feature is different from a historical non-motor vehicle feature;

the judging module 804 is further configured to judge that the driver has a vehicle-stealing behavior if it is detected that the non-motor vehicle features corresponding to the same driver features are different from the historical non-motor vehicle features.

Optionally, as shown in fig. 8, the clustering module 802 is further configured to input the non-motor vehicle feature into a clustering engine, and cluster to obtain history snapshot information based on the non-motor vehicle feature;

The detection module 803 is further configured to detect whether there is a situation in which the driver characteristics corresponding to the same non-motor vehicle are different from the historical driver characteristics;

the judging module 804 is further configured to judge that the vehicle theft behavior exists if it is detected that the driver characteristics corresponding to the same non-motor vehicle are different from the historical driver characteristics.

Optionally, as shown in fig. 9, the apparatus further includes:

A second obtaining module 807, configured to obtain, after determining that there is a vehicle theft behavior, peer information corresponding to a driver having the vehicle theft behavior, where the peer information includes peer times and peer information;

The map construction module 808 is configured to construct a relationship map about the driver with the vehicle theft behavior according to the peer information, where the relationship map includes peer times and peer information, and the relationship map is used to show peer relationships between the peer and the driver with the vehicle theft behavior.

Optionally, as shown in fig. 10, the apparatus further includes:

the comparison module 809 is configured to compare the peer personnel information with a forensic personnel library, where the forensic personnel library includes forensic information of a vehicle theft personnel;

the adjustment module 810 is configured to adjust the relationship map according to the comparison result, so as to obtain a suspected person relationship map with vehicle theft behavior, where the suspected person relationship map includes the number of times of the same traffic and corresponding information of front departments of vehicle theft personnel.

Optionally, as shown in fig. 11, the apparatus further includes:

A third obtaining module 811, configured to continuously take a snapshot of a driver with the vehicle theft behavior after determining that the vehicle theft behavior exists, and obtain corresponding snapshot information;

And the track module 812 is configured to form a track for the driver with the vehicle theft behavior according to the snapshot information, where the track is used for analyzing the place where the driver with the vehicle theft behavior appears.

Optionally, as shown in fig. 12, the first obtaining module 801 includes:

An acquiring unit 8011, configured to acquire non-motor vehicle motion information and personnel motion information in the tracking image, where the non-motor vehicle motion information includes non-motor vehicle features, and the personnel motion information includes driver features;

And the matching unit 8012 is configured to match the non-motor vehicle characteristic with a corresponding driver characteristic according to the non-motor vehicle motion information and the personnel motion information, so as to form a non-motor vehicle driving characteristic.

It should be noted that, the device for detecting the vehicle theft behavior of the non-motor vehicle provided by the embodiment of the invention can be applied to a device for detecting the vehicle theft behavior of the non-motor vehicle, for example: and equipment such as community cameras, traffic cameras, computers and servers for detecting the vehicle theft behavior of the non-motor vehicles.

The non-motor vehicle traffic violation monitoring device provided by the embodiment of the invention can realize each implementation mode in the method embodiments of fig. 1 to 8 and corresponding beneficial effects, and in order to avoid repetition, the description is omitted here.

Referring to fig. 13, fig. 13 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, as shown in fig. 13, including: a memory 1302, a processor 1301, and a computer program stored on and executable on the memory, wherein:

Processor 1301 is configured to invoke a computer program stored in memory 1302, performing the steps of:

Detecting whether the driving characteristics of the non-motor vehicle are different from the corresponding historical snapshot information;

Optionally, the inputting the non-motor vehicle driving feature into the clustering engine, where the clustering is performed by the processor 1301, obtains the history snapshot information based on the non-motor vehicle driving feature, where the history snapshot information includes:

inputting the driver characteristics into a clustering engine, and clustering to obtain history snapshot information based on the driver characteristics;

The detecting performed by the processor 1301 if there is a situation that the driving characteristics of the non-motor vehicle are different from the corresponding history snapshot information includes:

detecting whether the non-motor vehicle characteristics corresponding to the same driver characteristics are different from the historical non-motor vehicle characteristics;

The determining, by the processor 1301, whether there is a vehicle theft behavior according to the detection result includes:

if the non-motor vehicle features corresponding to the same driver features are detected to be different from the historical non-motor vehicle features, judging that the driver has the vehicle stealing behavior.

inputting the non-motor vehicle features into a clustering engine, and clustering to obtain history snapshot information based on the non-motor vehicle features;

Detecting whether the situation that the characteristics of the drivers corresponding to the same non-motor vehicle are different from the characteristics of the historic drivers exists or not;

If the fact that the characteristics of the drivers corresponding to the same non-motor vehicle are different from the characteristics of the historical drivers is detected, the vehicle theft behavior is judged.

Optionally, the processor 1301 further executes instructions that include:

After judging that the vehicle theft behavior exists, acquiring peer information corresponding to a driver with the vehicle theft behavior, wherein the peer information comprises peer times and peer information;

and constructing a relationship map of the drivers with the vehicle theft behaviors according to the peer information, wherein the relationship map comprises peer times and peer information, and the relationship map is used for displaying peer relationships between the peers and the drivers with the vehicle theft behaviors.

Optionally, the processor 1301 further executes instructions that include:

comparing the staff information with a front department staff library, wherein the front department staff library comprises front department information of vehicle theft staff;

And adjusting the relation map according to the comparison result to obtain a suspected person relation map with vehicle theft behaviors, wherein the suspected person relation map comprises the same times and corresponding vehicle theft personnel forensic information.

Optionally, the processor 1301 further executes instructions that include:

After judging that the vehicle theft behavior exists, continuously taking a snapshot of a driver with the vehicle theft behavior to acquire corresponding snapshot information;

And forming a track aiming at the driver with the vehicle theft behavior according to the snapshot information, wherein the track is used for analyzing the place where the driver with the vehicle theft behavior appears.

Optionally, the acquiring the non-motor vehicle driving feature in the tracking image performed by the processor 1301 includes:

Acquiring non-motor vehicle motion information and personnel motion information in a tracking image, wherein the non-motor vehicle motion information comprises non-motor vehicle characteristics, and the personnel motion information comprises driver characteristics;

And matching the non-motor vehicle characteristics with corresponding driver characteristics according to the non-motor vehicle movement information and the personnel movement information to form non-motor vehicle driving characteristics.

It should be noted that, the electronic device provided by the embodiment of the present invention may be applied to a device for detecting vehicle theft behavior of a non-motor vehicle, for example: and equipment such as community cameras, traffic cameras, computers and servers for detecting the vehicle theft behavior of the non-motor vehicles.

The electronic device provided by the embodiment of the present invention can implement each implementation manner and corresponding beneficial effects in the method embodiments of fig. 1 to 7, and in order to avoid repetition, a detailed description is omitted here.

The embodiment of the invention also provides a computer readable storage medium, and a computer program is stored on the computer readable storage medium, and when the computer program is executed by a processor, the processes of the non-motor vehicle traffic violation monitoring method embodiment provided by the embodiment of the invention are realized, and the same technical effects can be achieved, so that repetition is avoided, and the description is omitted here.

Those skilled in the art will appreciate that implementing all or part of the above-described methods in accordance with the embodiments may be accomplished by way of a computer program stored on a computer readable storage medium, which when executed may comprise the steps of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM) or the like.

The foregoing disclosure is illustrative of the present invention and is not to be construed as limiting the scope of the invention, which is defined by the appended claims.

Claims

1. A method for detecting vehicle theft behavior of a non-motor vehicle is characterized by comprising the following steps:

judging whether the vehicle theft behavior exists or not according to the detection result;

The detecting whether the driving characteristics of the non-motor vehicle are different from the corresponding historical snapshot information comprises the following steps:

Detecting whether the non-motor vehicle characteristics corresponding to the same driver characteristics are different from the historical non-motor vehicle characteristics; or (b)

And detecting whether the situation that the driver characteristics corresponding to the same non-motor vehicle are different from the historical driver characteristics exists.

2. The method of claim 1, wherein the inputting the non-motor vehicle drive feature into a clustering engine, clustering to obtain historical snap-shot information based on the non-motor vehicle drive feature, comprises:

Judging whether the vehicle theft behavior exists according to the detection result, comprising the following steps:

3. The method of claim 1, wherein the inputting the non-motor vehicle drive feature into a clustering engine, clustering to obtain historical snap-shot information based on the non-motor vehicle drive feature, comprises:

4. The method of claim 1, wherein the method further comprises:

5. The method of claim 4, wherein the method further comprises:

6. The method of claim 1, wherein the method further comprises:

7. The method of any one of claims 1 to 6, wherein the acquiring non-motor vehicle drive characteristics in the tracking image comprises:

8. A device for detecting vehicle theft behavior of a non-motor vehicle, comprising:

the detection module is also used for detecting whether the condition that the non-motor vehicle characteristics corresponding to the same driver characteristics are different from the historical non-motor vehicle characteristics exists or not;

The detection module is also used for detecting whether the situation that the characteristics of the drivers corresponding to the same non-motor vehicle are different from the characteristics of the historic drivers exists or not;

9. An electronic device, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps in the method of detecting vehicle theft behavior of a non-motor vehicle as claimed in any one of claims 1 to 7 when the computer program is executed.

10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of the method for detecting vehicle theft behavior of a non-motor vehicle as claimed in any one of claims 1 to 7.