Target object determination method and apparatus, storage medium, and electronic device

Abstract

The application discloses a target object determination method and device, a storage medium and an electronic device, and the method comprises the following steps: determining an image acquisition device corresponding to a first position where a target event occurs, and acquiring a plurality of pictures collected by the image acquisition device within a preset time period before a current time; performing image analysis on the plurality of pictures to obtain a plurality of target objects, wherein each target object in the plurality of target objects has one or more labels, and the label is used for indicating the behavior characteristics of the target object; for each target object in the plurality of target objects, the correlation coefficient of each target object is determined according to the one or more labels of each target object, and the correlation coefficients of the plurality of target objects are obtained; and the target object with the largest correlation with the target event is determined from the plurality of target objects according to the correlation coefficients of the plurality of target objects, wherein the correlation coefficient is used for indicating the correlation degree of the target object and the event.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of communication, in particular to a target object determination method and device, a storage medium and an electronic device. BACKGROUND

[0002] With the rapid development of society, more and more places have cameras, and the video pictures taken by these cameras can be called by the management department, and then if a target event occurs, the management personnel can manually reason according to the specific event situation combined with the pictures taken by the camera to reason out the person associated with the target event. But this scheme will consume a lot of time of the management personnel, and manual watching of video and pictures is easy to have loopholes.

[0003] For related technologies, the traditional method can only manually reason out the person associated with the target event, resulting in low reasoning efficiency. Currently, no effective solution has been proposed.

[0004] Therefore, it is necessary to improve the related technology to overcome the defects in the related technology. SUMMARY

[0005] The embodiments of the present application provide a target object determination method and device, a storage medium and an electronic device to at least solve the problem that the traditional method can only manually reason out the person associated with the target event, resulting in low reasoning efficiency.

[0006] According to an aspect of the embodiments of the present application, a target object determination method is provided, comprising: determining an image acquisition device corresponding to a first position where a target event occurs, and acquiring a plurality of pictures collected by the image acquisition device in a preset time period before the current time, wherein the collection area of the image acquisition device includes the first position; performing image analysis on the plurality of pictures to obtain a plurality of target objects, wherein each target object in the plurality of target objects has one or more labels, and the label is used to indicate the behavior characteristics of the target object; for each target object in the plurality of target objects, determining the correlation coefficient of the each target object according to the one or more labels of the each target object, to obtain the correlation coefficients of the plurality of target objects; and determining the target object with the largest correlation with the target event from the plurality of target objects according to the correlation coefficients of the plurality of target objects, wherein the correlation coefficient is used to indicate the correlation degree of the target object with the event.

[0007] Further, the image analysis on the plurality of pictures obtains a plurality of target objects, including: performing image analysis on the plurality of pictures to obtain a plurality of first target objects, wherein the first target objects have first labels, and the first labels are used to indicate current behavior characteristics of the first target objects; performing image analysis on the plurality of pictures to obtain a plurality of second target objects, wherein the second target objects have second labels, and the second labels are used to indicate historical behavior characteristics of the second target objects; and performing image analysis on the plurality of pictures to obtain a plurality of third target objects, wherein the third target objects have third labels, and the third labels are used to indicate behavior characteristics of the third target objects using vehicles; wherein the target objects include the first target objects, the second target objects, and the third target objects; and the labels include the first labels, the second labels, and the third labels.

[0008] Further, the image analysis on the plurality of pictures obtains a plurality of first target objects, including: inputting the plurality of pictures into a first neural network model to obtain a plurality of first target objects, wherein the first neural network model is used to determine the first target objects having the first labels from a plurality of objects recorded in the plurality of pictures.

[0009] Further, the image analysis on the plurality of pictures obtains a plurality of second target objects, including: inputting the plurality of pictures into a second neural network model to obtain all objects recorded in the plurality of pictures; and determining a plurality of second target objects recorded in a target file from the all objects, wherein the objects recorded in the target file have second labels, and the second labels are used to indicate historical behavior characteristics of the objects.

[0010] Further, the image analysis on the plurality of pictures obtains a plurality of third target objects, including: inputting the plurality of pictures into a third neural network model to obtain a plurality of target vehicles, wherein the target vehicles have vehicle labels, and the vehicle labels are used to indicate current behavior characteristics of the target vehicles; determining a third target object corresponding to each target vehicle in the plurality of target vehicles according to a license plate number of the each target vehicle to obtain a plurality of third target objects, wherein the third labels and the vehicle labels are in one-to-one correspondence.

[0011] Further, the relevance score of each target object is determined according to one or more labels of the target object, including: in the case that the target object has only one label, determining a weight corresponding to the label, and determining the weight as the relevance score of the target object; in the case that the target object has multiple labels, determining multiple weights corresponding to the multiple labels, and determining the sum of the multiple weights as the relevance score of the target object.

[0012] Further, the target object most relevant to the target event is determined from the multiple target objects according to the relevance scores of the multiple target objects, including: determining a target ranking of the relevance score of each target object in the multiple relevance scores of the multiple target objects; and determining the target object corresponding to a target ranking greater than a preset threshold as the target object most relevant to the target event.

[0013] Further, the image acquisition device corresponding to the first position where the target event occurs is determined, including: obtaining the first position and a radius distance set for the first position; determining a circular region with the first position as the center and the radius distance as the radius as a target region to be searched; and determining one or more image acquisition devices located in the target region as the image acquisition device corresponding to the first position.

[0014] According to another aspect of the embodiment of the present application, a target object determination device is further provided, including: a first processing module configured to determine an image acquisition device corresponding to a first position where a target event occurs, and obtain multiple pictures collected by the image acquisition device in a preset time period before a current time, wherein a collection region of the image acquisition device includes the first position; a second processing module configured to perform image analysis on the multiple pictures to obtain multiple target objects, wherein each target object in the multiple target objects has one or more labels, and the labels are used to indicate behavior characteristics of the target object; a first determination module configured to, for each target object in the multiple target objects, determine a relevance score of the target object according to one or more labels of the target object, and obtain relevance scores of the multiple target objects; and a second determination module configured to determine a target object most relevant to the target event from the multiple target objects according to the relevance scores of the multiple target objects, wherein the relevance score is used to indicate a degree of relevance of the target object to the event.

[0015] According to still another aspect of the embodiment of the present application, a computer readable storage medium is further provided, and the computer readable storage medium stores a computer program, wherein the computer program is configured to execute the target object determination method when running.

[0016] According to a further aspect of the embodiments of the present application, an electronic device is provided, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the above-mentioned target object determination method through the computer program.

[0017] Through the present application, the image acquisition device corresponding to the first position of the target event is determined, and a plurality of pictures acquired by the image acquisition device in a preset time period before the current time are obtained. Through image analysis on the plurality of pictures, a plurality of target objects are obtained, and the correlation coefficient of each target object is determined through one or more labels of each target object. Then, the target object most associated with the target event is determined according to the correlation coefficient. By using the above technical solution, the problem that the traditional method can only manually infer the person associated with the target event, resulting in low inference efficiency, is solved. Then, the collected pictures are analyzed using an algorithm to obtain the person associated with the target event, so that the inference efficiency is high. BRIEF DESCRIPTION OF DRAWINGS

[0018] The accompanying drawings, which are included to provide a further understanding of the present application and are incorporated in and constitute a part of this application, illustrate embodiments of the present application and serve to explain the principles of the present application, and do not limit the present application. In the drawings:

[0019] Figure 1 is a hardware structure block diagram of a computer terminal of the target object determination method according to the embodiments of the present application;

[0020] Figure 2 is a flowchart of the target object determination method according to the embodiments of the present application (one);

[0021] Figure 3 is a flowchart of the target object determination method according to the embodiments of the present application (two);

[0022] Figure 4 is a structure block diagram of the target object determination device according to the embodiments of the present application. DETAILED DESCRIPTION

[0023] In order for those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.

[0024] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0025] The methods and embodiments provided in this application can be executed on a computer terminal or similar computing device. Taking running on a computer terminal as an example, Figure 1 This is a hardware structure block diagram of a computer terminal for a method of determining the target object according to an embodiment of the present invention. For example... Figure 1 As shown, a computer terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor unit (MPU) or a programmable logic device (PLD)) and a memory 104 for storing data are also shown. In one exemplary embodiment, the computer terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the computer terminal described above. For example, the computer terminal may also include components that are more complex than those described above. Figure 1 The more or fewer components shown, or having the same Figure 1 Equivalent functions or ratios shown Figure 1 The functions shown have more different configurations.

[0026] The memory 104 can be used to store computer programs, such as software programs of application software and modules, such as a computer program corresponding to the target object determination method in the embodiments of the present application. The processor 102 performs various functional applications and data processing, i.e., implements the above method, by running the computer program stored in the memory 104. The memory 104 can include a high-speed random access memory, and can further include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include a memory remotely arranged with respect to the processor 102, which can be connected to the computer terminal through a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0027] The transmission device 106 is used to receive or send data via a network. Specific examples of the above network can include a wireless network provided by a communication provider of the computer terminal. In one example, the transmission device 106 includes a network adapter (NIC), which can be connected to other network devices through a base station so as to be able to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet in a wireless manner.

[0028] In order to solve the technical problems of the present application, an algorithm is provided, which associates multiple types of information such as event information, rule violation personnel information, and snapshot information, finds out the personnel who are most likely to have rule violation behaviors in a code manner by associating with events, and calculates their association degrees.

[0029] Specifically, in the present embodiment, a target object determination method is provided, Figure 2 is a flowchart of the target object determination method according to the embodiments of the present application, which includes the following steps:

[0030] In step S202, an image acquisition device corresponding to a first position where a target event occurs is determined, and multiple pictures collected by the image acquisition device within a preset time period before the current time are obtained, wherein the collection area of the image acquisition device includes the first position.

[0031] It should be noted that the target event includes but is not limited to an object loss event. The first position can be determined by a latitude and longitude manner, the image acquisition device includes but is not limited to a camera, and the preset time period can be three days before and after the occurrence time of the target event.

[0032] Step S204, image analysis is performed on the plurality of pictures to obtain a plurality of target objects, wherein each of the plurality of target objects has one or more labels, and the labels are used to indicate behavior characteristics of the target objects.

[0033] It should be noted that the target object is a person determined by analyzing the pictures. The behavior characteristics include, but are not limited to, early leaving and late returning, wandering, and having done a target event at a historical time.

[0034] Step S206, for each of the plurality of target objects, a correlation coefficient of each of the plurality of target objects is determined according to one or more labels of each of the plurality of target objects.

[0035] Step S208, a target object with the largest correlation with the target event is determined from the plurality of target objects according to the correlation coefficients of the plurality of target objects, wherein the correlation coefficients are used to indicate the degree of correlation between the target objects and the event.

[0036] Through the above steps, the image acquisition device corresponding to the first position where the target event occurs is determined, and a plurality of pictures collected by the image acquisition device in a preset time period before the current time are obtained. A plurality of target objects are obtained by image analysis on the plurality of pictures, and the correlation coefficient of each target object is determined by one or more labels of each target object. Then, the target object with the largest correlation with the target event is determined according to the correlation coefficient. By using the above technical solution, the problem that the traditional method can only manually infer the person associated with the target event, resulting in low inference efficiency, is solved. Then, the collected pictures are analyzed by using an algorithm to obtain the person associated with the target event, so that the inference efficiency is high.

[0037] In an exemplary embodiment, the image acquisition device corresponding to the first position where the target event occurs can be determined by the following method: obtaining the first position and a radius distance set for the first position; determining a circular region with the first position as the center and the radius distance as the radius as a target region to be searched; and determining one or more image acquisition devices with a geographical position in the target region as the image acquisition device corresponding to the first position.

[0038] For better understanding, the following is specifically described, assuming that the first position is represented by latitude and longitude, the first position is (longitude a, latitude b), and the preset radius is 1000 meters. A circle with a center of (longitude a, latitude b) and a radius of 1000 meters is determined as a target area, and one or more cameras in the target area are determined as image acquisition devices corresponding to the first position. Specifically, whether the camera is in the target area can be determined by the following method: calculating the distance from the camera to the first position, and determining the camera as the image acquisition device corresponding to the first position if the distance is less than or equal to 1000 meters. The distance calculation formula is as follows:

[0039] 6371000*acos(cos(radians(a.wd))*cos(radians(b.jd))*cos(radians(a.jd)-b.wd)+sin(radians(a.wd))*sin(radians(b.jd)));

[0040] wherein a.wd is the latitude of the first position, a.jd is the longitude of the first position, b.wd is the latitude of the camera, and b.jd is the longitude of the camera.

[0041] In an exemplary embodiment, image analysis is performed on the plurality of pictures to obtain a plurality of target objects, which can be achieved by the following method: performing image analysis on the plurality of pictures to obtain a plurality of first target objects, wherein the first target objects have a first label, and the first label is used to indicate the current behavior characteristics of the first target objects; and performing image analysis on the plurality of pictures to obtain a plurality of second target objects, wherein the second target objects have a second label, and the second label is used to indicate the historical behavior characteristics of the second target objects; and performing image analysis on the plurality of pictures to obtain a plurality of third target objects, wherein the third target objects have a third label, and the third label is used to indicate the behavior characteristics of the third target objects using vehicles; wherein the target objects include the first target objects, the second target objects, and the third target objects; and the labels include the first label, the second label, and the third label.

[0042] It should be noted that the current behavior characteristics of the target object include but are not limited to: early out and late in, nocturnal, etc., and the historical behavior characteristics of the target object include but are not limited to: no fixed residence, having done the target event at a historical time, etc. The behavior characteristics of the target object using the vehicle include but are not limited to using a fake license plate, early out and late in by car, etc. Since there are multiple labels, the target object with different labels can be determined by different algorithms. That is, a first target object with an early out and late in label is determined, a second target object with a no fixed residence label is determined, etc. It should be noted that the first target object and the second target object can be the same. That is, there is a target object that has both the early out and late in label and the no fixed residence label.

[0043] For better understanding, in an exemplary embodiment, the image analysis of the plurality of pictures to obtain a plurality of first target objects can be achieved by: inputting the plurality of pictures into a first neural network model to obtain a plurality of first target objects, wherein the first neural network model is used to determine the first target objects with the first label from a plurality of objects recorded in the plurality of pictures.

[0044] That is, a neural network model can be trained in advance by data, so that the neural network model can determine the first target object with the first label in the picture according to the characteristics of the object in the picture. Assuming that the camera has taken 10 pictures, and then the 10 pictures are input into the first neural network, so that the first neural network can determine 2 first target objects with early out and late in (first label).

[0045] In an exemplary embodiment, the image analysis of the plurality of pictures to obtain a plurality of second target objects can be achieved by: inputting the plurality of pictures into a second neural network model to obtain all objects recorded in the plurality of pictures; determining a plurality of second target objects recorded in a target file from the all objects, wherein the objects recorded in the target file have a second label, and the second label is used to indicate the historical behavior characteristics of the object.

[0046] The object is a person existing in the picture, and the identity of all persons existing in the plurality of pictures taken by the image acquisition device can be determined by the second neural network. Assuming that the target event is an object loss event, the target file can be obtained from the management system, and the target file records persons having corresponding violation behaviors at a historical time, and whether the person in the picture is recorded in the target file can be determined. If it is recorded in the target file, the person has the second label.

[0047] In an example embodiment, the image analysis on the plurality of pictures to obtain a plurality of third target objects can be achieved by inputting the plurality of pictures into a third neural network model to obtain a plurality of target vehicles, wherein the target vehicle has a vehicle label indicating a current behavior feature of the target vehicle; determining a third target object corresponding to each target vehicle in the plurality of target vehicles according to a license plate number of the each target vehicle to obtain a plurality of third target objects, wherein the third label corresponds to the vehicle label one by one.

[0048] That is, in the plurality of pictures taken by the image acquisition device, not only people exist, but also vehicles exist, and then the vehicles recorded in the pictures can be analyzed to determine vehicles with vehicle labels, which include but are not limited to early-out late-in vehicles, day-out night-in vehicles, and fake-plate vehicles. After determining the target vehicle, the owner (third target object) of the vehicle can be determined according to the license plate number of the target vehicle. It should be noted that the third label of the third target object corresponds to the vehicle label one by one. For example, the vehicle label of vehicle A is an early-out late-in vehicle, and the third label of the third target object corresponding to vehicle A is an early-out late-in vehicle.

[0049] In an example embodiment, the correlation coefficient of each target object is determined according to one or more labels of the each target object, which can be determined by the following method: in the case that the each target object has only one label, determining a weight corresponding to the label, and determining the weight as the correlation coefficient of the each target object. In the case that the each target object has a plurality of labels, determining a plurality of weights corresponding to the plurality of labels, and determining the sum of the plurality of weights as the correlation coefficient of the each target object.

[0050] That is, if the target object has only one early-out late-in label, the weight of the early-out late-in label of the target event is determined, which is assumed to be 40, and the weight 40 is determined as the correlation coefficient of the target object and the target event. If the target object has a plurality of labels, for example, an early-out late-in label, a no-fixed residence label, and a fake-plate vehicle label, the weights of the three labels are determined, which are assumed to be 40, 10, and 10, and the sum of the three weights 60 is determined as the correlation coefficient of the target object and the target event.

[0051] In an example embodiment, the target object most associated with the target event is determined from the plurality of target objects according to the correlation coefficients of the plurality of target objects, which can be achieved by the following method: determining a target ranking of the correlation coefficient of each target object in the plurality of correlation coefficients of the plurality of target objects; determining the target object corresponding to a target ranking greater than a preset threshold as the target object most associated with the target event.

[0052] It should be noted that if the correlation coefficient of a target object is larger, it indicates that the correlation degree of the target object and the target event is higher, and then the correlation coefficient of each target object is ranked in all correlation coefficients, and the multiple target objects ranked higher than the fifth (preset threshold) are determined as the target objects most correlated with the target event.

[0053] Obviously, the above-described embodiments are only a part of the embodiments of the present application, not all. In order to better understand the above-mentioned target object determination method, the above process is described below in combination with the embodiments, but not used to limit the technical solutions of the embodiments of the present application, specifically:

[0054] In an optional embodiment, the possible rule violation behavior persons (equivalent to the target objects in the above embodiment) of various labels are associated by the event occurrence place snapshot data, the personnel information associated is summed by the correlation ratio algorithm to obtain the correlation ratio of each possible rule violation behavior person, and then the descending order is arranged to find the possible rule violation behavior person with the largest correlation ratio related to the event.

[0055] Specifically, Figure 3 is a flowchart of the target object determination method according to the embodiments of the present application (two), comprising:

[0056] S1: colliding the event table and the device table;

[0057] It should be noted that the event table records the name of the event, the type of the event, the time of the event occurrence, the location of the event occurrence, etc. The device table records the number of the camera, the location of the camera, etc. Then, in the case of determining the target area according to the preset distance, the event table and the device table can be collided, the camera in the target area is located according to the latitude and longitude, and a table with the number of multiple cameras recorded is obtained.

[0058] For example: the object loss event and the device table are collided, the range is located within 1000 meters, and the latitude and longitude in the two tables are located according to the distance calculation formula: 6371000*acos(cos(radians(a.wd))*cos(radians(b.jd))*cos(radians(a.jd)-b.wd)+sin(radians(a.wd))*sin(radians(b.jd))), and then the camera in the target area in the device table is determined.

[0059] S2: colliding the result table formed by S1 and the lens snapshot table;

[0060] Specifically, it needs to be explained that the card mouth snapshot table records the pictures snapped by the camera. The result table of S1 and the card mouth personnel snapshot information table are collided, and then the snapshot data of the involved card mouth can be found. It needs to be explained that the snapshot data of the camera determined in S2 in the three days before and after the event occurrence time can be obtained, so that the personnel who may have violated the rules can be found as much as possible.

[0061] S3: Collide each associated label data with the result of S2;

[0062] The result table of S2, i.e. the face trajectory snapshot picture screening and each associated label data, are collided to find the trajectory snapshot of the personnel who may have violated the rules. The labels include: personnel loitering, personnel without valid address, personnel day out and night in, personnel early out and late in, etc.

[0063] For example: Collide the personnel without valid address with the face trajectory snapshot picture table to find the face card mouth snapshot record of the personnel without valid address in the three days before the event occurrence time to the three days after the event occurrence time in the card mouth around the event occurrence place, i.e. find the target object with associated label.

[0064] S4: The operation of S2 and S3 is also performed on the vehicle label;

[0065] The result table of S1 and the card mouth vehicle snapshot information table are collided to find the snapshot of the involved card mouth, wherein the event range is set in the three days before and after the event occurrence time, so that the associated vehicle can be found as much as possible to form a vehicle trajectory snapshot table around the event occurrence place, and then collide with each associated vehicle label table to find the snapshot of the associated vehicle around the event occurrence place. The associated vehicle label includes: vehicle early out and late in, vehicle loitering, vehicle day out and night in, and fake license plate car, etc.

[0066] For example: The fake license plate car label data and the vehicle trajectory snapshot table around the event occurrence place are collided to find the vehicle snapshot of the fake license plate car in the three days before and after the event occurrence time around the event occurrence place.

[0067] It needs to be explained that after the vehicle snapshot is found, the vehicle owner table is collided to find the owner of the associated vehicle, and the vehicle owner information needs to be used subsequently.

[0068] S5: Merge the label vehicle owner and the personnel label;

[0069] The vehicle owner in S4 and the personnel label are merged, and the information such as the person's certificate number and license plate number is merged to form a new table, which provides a data basis for the subsequent association degree algorithm of the personnel who may have violated the rules.

[0070] S6: Perform row-to-column conversion on the result of S5;

[0071] The S5 people and vehicle combination table is converted from row to column, the row of the people and vehicle combination table is converted to the column of the people and vehicle combination table, and the existing label is marked with a value 1, and the label without value 0. For example:

[0072] Table 1

[0073] Event Name ID Number Tag A Zhang San 001 Person loitering A Zhang San 001 Daytime hiding at night

[0074] After the row-column conversion of Table 1, Table 2 can be obtained:

[0075] Table 2

[0076] Event Name ID Number Person loitering Daytime hiding at night A Zhang San 001 1 1

[0077] S7: The label integration rule of the table after the row-column conversion is performed;

[0078] According to the algorithm rule of the correlation degree in S6, each label is assigned a coefficient, the coefficient is a self-defined correlation coefficient (equivalent to the weight in the above embodiment), and then the label data is multiplied, and then the integration of the same person is summed, the correlation ratio of the person who may exist in the illegal behavior can be calculated (equivalent to the correlation coefficient in the above embodiment), and then descending order is arranged, so that the maximum value of the degree of correlation between the person and the event can be seen.

[0079] The correlation ratio algorithm formula is:

[0080] Correlation ratio = personnel lingering stay *

[10] + personnel day out *

[20] + suit card car *

[40] +...;

[0081] Wherein,

[10] ,

[20] , 40] are correlation coefficients, which are self-defined coefficients, and are reasonably distributed according to the nature of the label and the event, for example: the coefficient of the suit card car is reasonably enlarged in the vehicle loss event.

[0082] Finally, the result of the correlation ratio algorithm is:

[0083] Table 3

[0084] Event Name ID Number Person loitering Daytime hiding at night Degree of association A Zhang San 001 1 1 30

[0085] In addition, the above technical solutions of the embodiment of the application can find the correlation between the person who may exist in the illegal behavior and the event more intuitively by calculating the correlation ratio. Through multi-dimensional data fusion, the data and the event are integrated from the perspective of the vehicle and the person, so as to provide more comprehensive data and more ideas for the management personnel to solve the problem.

[0086] Those skilled in the art can clearly understand that the method according to the above-mentioned embodiments can be realized by means of software and necessary general hardware platforms, of course, can also be realized by hardware, but in many cases, the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of software products, which are stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and include a plurality of instructions for making a terminal device (which can be a mobile phone, computer, server, or network device, etc.) execute the method of each embodiment of the present application.

[0087] In the embodiment, a target object determining apparatus is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and will not be described again. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware, or a combination of software and hardware is also possible and is contemplated.

[0088] Figure 4 is a structural block diagram of a target object determining apparatus according to an embodiment of the present application, which comprises:

[0089] A first processing module 42 is configured to determine an image acquisition device corresponding to a first position where a target event occurs, and acquire a plurality of pictures collected by the image acquisition device in a preset time period before a current time, wherein a collection area of the image acquisition device includes the first position.

[0090] A second processing module 44 is configured to perform image analysis on the plurality of pictures to obtain a plurality of target objects, wherein each target object in the plurality of target objects has one or more labels, and the labels are used to indicate behavior characteristics of the target object.

[0091] A first determining module 46 is configured to determine a correlation coefficient of each target object in the plurality of target objects according to the one or more labels of the each target object, to obtain correlation coefficients of the plurality of target objects.

[0092] A second determining module 48 is configured to determine a target object with the largest correlation with a target event from the plurality of target objects according to the correlation coefficients of the plurality of target objects, wherein the correlation coefficient is used to indicate a correlation degree of the target object with the event.

[0093] The first position corresponding to the image acquisition device where the target event occurs is determined through the above module, and a plurality of pictures collected by the image acquisition device in a preset time period before the current time are obtained. A plurality of target objects are obtained by image analysis on the plurality of pictures, and the correlation coefficient of each target object is determined through one or more labels of each target object. Then, the target object most associated with the target event is determined according to the correlation coefficient. The above technical solution solves the problem that the traditional method can only manually infer the person associated with the target event, resulting in low inference efficiency. Then, the collected pictures are analyzed in an algorithmic manner to obtain the person associated with the target event, so that the inference efficiency is high.

[0094] Optionally, the second processing module 44 is further configured to perform image analysis on the plurality of pictures to obtain a plurality of first target objects, wherein the first target objects have first labels, and the first labels are used to indicate current behavior characteristics of the first target objects; perform image analysis on the plurality of pictures to obtain a plurality of second target objects, wherein the second target objects have second labels, and the second labels are used to indicate historical behavior characteristics of the second target objects; and perform image analysis on the plurality of pictures to obtain a plurality of third target objects, wherein the third target objects have third labels, and the third labels are used to indicate behavior characteristics of the third target objects using vehicles; wherein the target objects include the first target objects, the second target objects, and the third target objects; and the labels include the first labels, the second labels, and the third labels.

[0095] Optionally, the second processing module 44 is further configured to input the plurality of pictures into a first neural network model to obtain a plurality of first target objects, wherein the first neural network model is used to determine the first target objects having the first labels from a plurality of objects recorded in the plurality of pictures.

[0096] Optionally, the second processing module 44 is further configured to input the plurality of pictures into a second neural network model to obtain all objects recorded in the plurality of pictures; and determine a plurality of second target objects recorded in a target file from the all objects, wherein the objects recorded in the target file have second labels, and the second labels are used to indicate historical behavior characteristics of the objects.

[0097] Optionally, the second processing module 44 is further configured to input the plurality of pictures into a third neural network model to obtain a plurality of target vehicles, wherein the target vehicles have vehicle labels, the vehicle labels are used to indicate current behavior characteristics of the target vehicles, and a third target object corresponding to each target vehicle in the plurality of target vehicles is determined according to a license plate number of the each target vehicle to obtain a plurality of third target objects, wherein the third labels correspond to the vehicle labels one by one.

[0098] Optionally, the first determining module 46 is further configured to, in a case where each target object has only one label, determine a weight corresponding to the label, and determine the weight as a correlation coefficient of the each target object, and in a case where each target object has a plurality of labels, determine a plurality of weights corresponding to the plurality of labels, and determine a sum of the plurality of weights as the correlation coefficient of the each target object.

[0099] Optionally, the second determining module 48 is further configured to determine a target ranking of the correlation coefficient of each target object in the plurality of target objects in the plurality of correlation coefficients of the plurality of target objects, and determine a target object corresponding to a target ranking greater than a preset threshold as a target object most closely associated with a target event.

[0100] Optionally, the first processing module 42 is further configured to obtain the first position and a radius distance set for the first position, determine a circular region with the first position as a center and the radius distance as a radius as a target region to be searched, and determine one or more image acquisition devices with a geographical position located in the target region as image acquisition devices corresponding to the first position.

[0101] Embodiments of the present application also provide a computer readable storage medium having a computer program stored therein, wherein the computer program is configured to execute the steps in any of the method embodiments described above when running.

[0102] Optionally, in the present embodiment, the storage medium can be configured to store a computer program for executing the following steps:

[0103] S1, determining image acquisition devices corresponding to a first position where a target event occurs, and obtaining a plurality of pictures collected by the image acquisition devices in a preset time period before a current time, wherein a collection region of the image acquisition devices includes the first position;

[0104] S2, performing image analysis on the plurality of pictures to obtain a plurality of target objects, wherein each target object in the plurality of target objects has one or more labels, and the labels are used to indicate behavior characteristics of the target objects;

[0105] S3, for each target object in the plurality of target objects, determining a correlation coefficient of the each target object according to one or more labels possessed by the each target object, to obtain correlation coefficients of the plurality of target objects;

[0106] S4, determining a target object with the largest correlation with the target event from the plurality of target objects according to the correlation coefficients of the plurality of target objects, wherein the correlation coefficient is used to indicate a degree of correlation of the target object with the event.

[0107] In an example embodiment, the computer readable storage medium described above can include, but is not limited to, a U disk, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store computer programs.

[0108] The specific examples in the embodiment can refer to the examples described in the above embodiments and example implementations, and the embodiment will not be described here.

[0109] The embodiment of the application further provides an electronic device, including a memory and a processor, the memory stores a computer program, and the processor is configured to execute the computer program to perform the steps in any of the above method embodiments.

[0110] Optionally, in the embodiment, the processor can be configured to execute the following steps by the computer program:

[0111] S1, determining an image acquisition device corresponding to a first position where a target event occurs, and acquiring a plurality of pictures collected by the image acquisition device in a preset time period before a current time, wherein a collection area of the image acquisition device includes the first position;

[0112] S2, performing image analysis on the plurality of pictures to obtain a plurality of target objects, wherein each target object in the plurality of target objects has one or more labels, and the label is used to indicate a behavior feature of the target object;

[0113] S3, for each target object in the plurality of target objects, determining a correlation coefficient of the each target object according to one or more labels possessed by the each target object, to obtain correlation coefficients of the plurality of target objects;

[0114] S4, determining a target object with the largest correlation with the target event from the plurality of target objects according to the correlation coefficients of the plurality of target objects, wherein the correlation coefficient is used to indicate a degree of correlation of the target object with the event.

[0115] In one example embodiment, the electronic device described above can further include a transmission device connected to the processor and an input / output device connected to the processor.

[0116] The specific examples in the present embodiment can refer to the examples described in the above embodiments and exemplary implementations, which will not be repeated here.

[0117] Obviously, those skilled in the art should understand that the modules or steps of the present application described above can be realized by general computing devices, which can be centralized on a single computing device or distributed on a network composed of multiple computing devices, which can be realized by program codes executable by the computing devices, so that they can be stored in storage devices and executed by the computing devices, and in some cases, the steps shown or described can be executed in different order, or they can be respectively manufactured into individual integrated circuit modules, or multiple modules or steps among them can be manufactured into a single integrated circuit module to realize. Thus, the present application is not limited to any specific combination of hardware and software.

[0118] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for determining a target object, characterized in that, include: The image acquisition device corresponding to the first location where the target event occurs is determined, and multiple images acquired by the image acquisition device within a preset time period before the current time are obtained, wherein the acquisition area of ​​the image acquisition device includes the first location; Image analysis is performed on the multiple images to obtain multiple target objects, wherein each of the multiple target objects has one or more tags, and the tags are used to indicate the behavioral characteristics of the target object; For each of the plurality of target objects, the association coefficient of each target object is determined based on one or more labels that each target object has, thereby obtaining the association coefficient of the plurality of target objects; Based on the correlation coefficients of the multiple target objects, the target object that is most associated with the target event is determined from the multiple target objects, wherein the correlation coefficients are used to indicate the degree of association between the target object and the event; The method involves image analysis of the multiple images to obtain multiple second target objects from the multiple target objects: inputting the multiple images into a second neural network model to obtain all objects recorded in the multiple images; determining multiple second target objects recorded in a target file from all objects, wherein the objects recorded in the target file have a second label, the second label being used to indicate the historical behavioral characteristics of the object; the target object includes: the second target object; the label includes: the second label.

2. The method according to claim 1, characterized in that, Image analysis is performed on the multiple images to obtain multiple target objects, including: Image analysis is performed on the multiple images to obtain multiple first target objects, wherein each first target object has a first label, and the first label is used to indicate the current behavioral characteristics of the first target object; Image analysis is performed on the multiple images to obtain multiple third target objects, wherein each third target object has a third label, which is used to indicate the behavioral characteristics of the third target object using a vehicle; The target object further includes: the first target object and the third target object; the tag further includes: the first tag and the third tag.

3. The method according to claim 2, characterized in that, Image analysis is performed on the multiple images to obtain multiple first target objects, including: The multiple images are input into a first neural network model to obtain multiple first target objects, wherein the first neural network model is used to determine the first target object with the first label from the multiple objects recorded in the multiple images.

4. The method according to claim 2, characterized in that, Image analysis was performed on the multiple images to obtain multiple third target objects, including: The multiple images are input into a third neural network model to obtain multiple target vehicles, wherein each target vehicle has a vehicle label, which is used to indicate the current behavioral characteristics of the target vehicle; Based on the license plate number of each of the multiple target vehicles, a third target object corresponding to each target vehicle is determined, resulting in multiple third target objects, wherein the third tag corresponds one-to-one with the vehicle tag.

5. The method according to claim 1, characterized in that, Determining the association coefficient of each target object based on one or more tags possessed by each target object includes: In the case that each target object has only one label, the weight corresponding to the label is determined, and the weight is determined as the association coefficient of each target object; When each target object has multiple labels, multiple weights corresponding to the multiple labels are determined, and the sum of the multiple weights is determined as the correlation coefficient of each target object.

6. The method according to claim 1, characterized in that, Based on the correlation coefficients of the multiple target objects, the target object most strongly associated with the target event is determined from among the multiple target objects, including: Determine the target ranking of the correlation coefficient of each target object among the multiple target objects; The target objects whose ranking is greater than a preset threshold are identified as the target objects most closely associated with the target event.

7. The method for determining a target object according to claim 1, characterized in that, The image acquisition device that determines the first location where the target event occurs includes: Obtain the first position and the radius distance set for the first position; The circular region centered at the first position and with the radius distance as its radius is defined as the target region to be searched. One or more image acquisition devices located in the target area are identified as the image acquisition devices corresponding to the first location.

8. A device for determining a target object, characterized in that, include: The first processing module is used to determine the image acquisition device corresponding to the first location where the target event occurs, and to acquire multiple images acquired by the image acquisition device within a preset time period before the current time, wherein the acquisition area of ​​the image acquisition device includes the first location. The second processing module is used to perform image analysis on the multiple images to obtain multiple target objects, wherein each of the multiple target objects has one or more tags, and the tags are used to indicate the behavioral characteristics of the target object; The first determining module is used to determine the association coefficient of each target object based on one or more labels of each target object among the plurality of target objects, thereby obtaining the association coefficient of the plurality of target objects; The second determining module is used to determine the target object that is most associated with the target event from the plurality of target objects based on the correlation coefficient of the plurality of target objects, wherein the correlation coefficient is used to indicate the degree of association between the target object and the event; The second processing module is further configured to perform image analysis on the multiple images to obtain multiple second target objects among the multiple target objects by: inputting the multiple images into a second neural network model to obtain all objects recorded in the multiple images; determining multiple second target objects recorded in a target file from all objects, wherein the objects recorded in the target file have a second label, the second label being used to indicate the historical behavioral characteristics of the object; the target object includes: the second target object; the label includes: the second label.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein the program, when executed, performs the method described in any one of claims 1 to 7.

10. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to execute the method described in any one of claims 1 to 7 through the computer program.

Images