CN111435435A

CN111435435A - Method, device, server and system for identifying pedestrians

Info

Publication number: CN111435435A
Application number: CN201911259493.XA
Authority: CN
Inventors: 陈钦; 丁玲德; 蒋庆萍; 周伟军
Original assignee: Hangzhou Hikvision Digital Technology Co Ltd
Current assignee: Hangzhou Hikvision Digital Technology Co Ltd
Priority date: 2019-12-10
Filing date: 2019-12-10
Publication date: 2020-07-21
Also published as: WO2021114985A1

Abstract

The embodiment of the invention provides a method, a device, a server and a system for identifying pedestrians, wherein the method comprises the steps of analyzing an acquired pedestrian image, obtaining the snapshot time and the snapshot place of each pedestrian in the pedestrian image, determining that at least two pedestrians with the distance between the snapshot places within a preset range and the time interval of the snapshot time within a first preset time interval are the same pedestrian, generating a same pedestrian record once, locally recording the same pedestrian record among the pedestrians without identifying the pedestrians, providing query basis for a query staff, outputting the locally recorded same pedestrian record of each pedestrian instead of outputting only the same pedestrian record of a specific pedestrian, and meeting increasingly complex pedestrian analysis requirements.

Description

Method, device, server and system for identifying pedestrians

Technical Field

The invention relates to the technical field of security monitoring, in particular to a method, a device, a server and a system for identifying a co-pedestrian.

Background

The identification of the pedestrians is a key technology in the field of security protection, and is used for identifying whether two pedestrians have a same-row relationship, so that on one hand, the social relationship between the pedestrians can be cleared, potential risks can be found, and prevention can be timely made, and on the other hand, when the risks occur, information of the pedestrians, information of time, places and the like of the risks can be acquired, and important clues are provided for detection.

Since some designated pedestrians (such as suspects, criminals, etc.) need to be monitored in general scenes, the information of the co-pedestrians of the designated pedestrians needs to be obtained. In the corresponding method for identifying the co-pedestrian, after a plurality of pedestrian images acquired by monitoring equipment are acquired, the specified pedestrian in each pedestrian image is identified firstly, whether other pedestrians and the specified pedestrian appear in one pedestrian image continuously or not is judged, and if any pedestrian appears in one pedestrian image continuously and the specified pedestrian appears in one pedestrian image, the pedestrian and the specified pedestrian can be determined to be in the co-pedestrian relationship. Therefore, when an inquiry request is made, the inquiry personnel inputs the pedestrian information of the appointed pedestrian, so that other pedestrians having the same-row relation with the appointed pedestrian can be known, and an analysis basis is provided for behavior monitoring of the appointed pedestrian.

However, with the continuous development of big data, the scenes of analysis of pedestrians are more and more complex, and what the inquirer needs to look over is not limited to pedestrians related to the specified pedestrian, but the requirement of analysis of pedestrians that is more and more complex cannot be met because only the pedestrians of the specified pedestrian are recorded.

Disclosure of Invention

The embodiment of the invention aims to provide a method, a device, a server and a system for identifying pedestrians so as to meet increasingly complex pedestrian analysis requirements. The specific technical scheme is as follows:

in a first aspect, an embodiment of the present invention provides a method for identifying a peer, where the method includes:

acquiring a pedestrian image acquired by monitoring equipment;

analyzing the pedestrian images, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian images;

and if the distance between the snapshot places of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a co-pedestrian record between the at least two pedestrians once.

Optionally, before the step of generating a co-pedestrian record between at least two pedestrians once if the distance between the snapshot locations of at least two pedestrians is within the preset range and the time interval of the snapshot times of the at least two pedestrians is within the first preset time interval, the method further includes:

identifying pedestrian characteristics of each pedestrian;

clustering the pedestrians on the basis of the pedestrian characteristics of the pedestrians;

if the distance between the snapshot places of at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a pedestrian-sharing record between the at least two pedestrians, including:

for at least two pedestrians belonging to different classes, if the distance between the snapshot places of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians once.

Optionally, the step of obtaining the pedestrian image collected by the monitoring device includes:

acquiring at least one pedestrian image acquired by monitoring equipment; or acquiring a plurality of pedestrian images acquired by a plurality of monitoring devices;

analyzing the pedestrian image, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian image, wherein the steps comprise:

analyzing a pedestrian image acquired by monitoring equipment, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian image, wherein the snapshot time is a timestamp of the monitoring equipment for acquiring the pedestrian image, and the snapshot place is the installation position of the monitoring equipment;

alternatively, the first and second electrodes may be,

analyzing a plurality of pedestrian images acquired by a monitoring device, and determining the snapshot time and the snapshot place of each pedestrian in each pedestrian image, wherein the snapshot time is a timestamp of the monitoring device for acquiring each pedestrian image, and the snapshot place is the installation position of the monitoring device;

alternatively, the first and second electrodes may be,

analyzing a plurality of pedestrian images acquired by a plurality of monitoring devices, and determining the snapshot time and the snapshot place of each pedestrian in each pedestrian image, wherein the snapshot time is a timestamp of each monitoring device for acquiring each pedestrian image, and the snapshot place is the installation position of each monitoring device for acquiring each pedestrian image;

generating a co-pedestrian record between at least two pedestrians once aiming at least two pedestrians in one pedestrian image acquired by one monitoring device;

alternatively, the first and second electrodes may be,

if the time interval of the snapshot time of at least two pedestrians in the multiple pedestrian images acquired by one monitoring device is within a first preset time interval, generating a same-pedestrian record of the at least two pedestrians at one time;

alternatively, the first and second electrodes may be,

if the distance between the snapshot places of at least two pedestrians in the multiple pedestrian images acquired by the multiple monitoring devices is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians once.

Optionally, after the step of analyzing the pedestrian image and determining the snapshot time and the snapshot location of each pedestrian in the pedestrian image, the method further includes:

storing the snapshot time and the snapshot place of each pedestrian to a first database;

extracting the snapshot time and the snapshot place of at least two pedestrians from the first database;

if the distance between the snapshot places of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians once;

and storing the co-current person record to a second database.

Optionally, if the distance between the snapshot locations of the at least two pedestrians is within the preset range and the time interval of the snapshot time of the at least two pedestrians is within the first preset time interval, the step of generating a record of the same pedestrian between the at least two pedestrians includes:

identifying at least two pedestrians of which the distance between the snapshot places is within a preset range and the time interval of the snapshot time is within a first preset time interval;

counting the times of continuously generating the co-pedestrian records between the at least two pedestrians in a second preset time interval;

if the times are larger than or equal to a first preset threshold value, keeping a one-time same-person record generated in a second preset time interval.

Optionally, after the step of generating a co-pedestrian record between at least two pedestrians once if the distance between the snapshot locations of at least two pedestrians is within the preset range and the time interval of the snapshot times of the at least two pedestrians is within the first preset time interval, the method further includes:

acquiring a co-pedestrian query request, wherein the co-pedestrian query request comprises a target pedestrian to be queried;

searching all co-pedestrian records in which the target pedestrian appears, and counting the number of occurrences of each pedestrian, which is the co-pedestrian with the target pedestrian, in each co-pedestrian record;

and outputting the records of the same pedestrians of which the occurrence times are greater than a second preset threshold value.

In a second aspect, an embodiment of the present invention provides a method for identifying a peer, where the method includes:

searching all co-pedestrian records with target pedestrians from a database, and counting the occurrence times of the co-pedestrians who are co-pedestrians with the target pedestrians in each co-pedestrian record, wherein the co-pedestrian records of at least two pedestrians, which enable the interval of the snapshot places in the pedestrian images collected by the monitoring equipment to be within a preset range and the time interval of the snapshot time to be within a first preset time interval, are stored in the database;

and outputting the same-row record with the occurrence times larger than a preset threshold value.

In a third aspect, an embodiment of the present invention provides a device for identifying a fellow pedestrian, including:

the acquisition module is used for acquiring a pedestrian image acquired by the monitoring equipment;

the analysis module is used for analyzing the pedestrian images and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian images;

the identification recording module is used for generating a same-pedestrian record between at least two pedestrians once if the distance between the snapshot places of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval.

Optionally, the apparatus further comprises:

the clustering module is used for identifying the pedestrian characteristics of all pedestrians; clustering the pedestrians on the basis of the pedestrian characteristics of the pedestrians;

the identification recording module is specifically used for:

Optionally, the obtaining module is specifically configured to:

an analysis module specifically configured to:

alternatively, the first and second electrodes may be,

the identification recording module is specifically used for:

alternatively, the first and second electrodes may be,

if the time interval of the snapshot time of at least two pedestrians in the multiple pedestrian images acquired by one monitoring device is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians at one time;

alternatively, the first and second electrodes may be,

Optionally, the apparatus further comprises:

the storage module is used for storing the snapshot time and the snapshot place of each pedestrian to a first database;

the identification recording module is specifically used for:

and storing the co-current person record to a second database.

Optionally, the identification recording module is specifically configured to:

Optionally, the obtaining module is further configured to:

acquiring a co-pedestrian query request, wherein the co-pedestrian query request comprises pedestrian information of a target pedestrian to be queried;

the device also includes:

the searching module is used for searching all co-pedestrian records in which the target pedestrian appears according to the pedestrian information of the target pedestrian, and counting the number of times of occurrence of each pedestrian which is the co-pedestrian with the target pedestrian in each co-pedestrian record;

and the output module is used for outputting the records of the same pedestrians of which the occurrence times are greater than a second preset threshold value.

In a fourth aspect, an embodiment of the present invention provides a device for identifying a peer, where the device includes:

the system comprises an acquisition module, a query module and a query module, wherein the acquisition module is used for acquiring a co-pedestrian query request, and the co-pedestrian query request comprises a target pedestrian to be queried;

the system comprises a searching module, a judging module and a monitoring module, wherein the searching module is used for searching all co-pedestrian records with target pedestrians from a database and counting the occurrence times of the co-pedestrians, which are co-pedestrians with the target pedestrians, in each co-pedestrian record, wherein the co-pedestrian records of at least two pedestrians, which enable the interval of snapshot places in pedestrian images acquired by monitoring equipment to be within a preset range and the time interval of snapshot time to be within a first preset time interval, are stored in the database;

and the output module is used for outputting the same-row record with the occurrence frequency larger than the preset threshold value.

In a fifth aspect, an embodiment of the present invention provides a server, including a processor and a memory, where the memory is used for storing a computer program;

a processor, configured to implement the method provided by the first aspect or the method provided by the second aspect of the embodiments of the present invention when executing the computer program stored in the memory.

In a sixth aspect, an embodiment of the present invention provides a peer identification system, where the system includes multiple monitoring devices and a server;

the monitoring equipment is used for acquiring pedestrian images;

the server is used for acquiring a pedestrian image acquired by the monitoring equipment; analyzing the pedestrian images, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian images; and if the distance between the snapshot places of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a co-pedestrian record between the at least two pedestrians once.

Optionally, the system further comprises a first database and a second database;

the first database is used for storing snapshot time and snapshot places of all pedestrians obtained by analyzing the pedestrian images by the server;

and the second database is used for storing the co-pedestrian records generated by the server.

Optionally, the system further comprises a client;

the server is also used for acquiring a co-pedestrian query request, wherein the co-pedestrian query request comprises pedestrian information of a target pedestrian to be queried; searching all co-pedestrian records in which the target pedestrian appears according to the pedestrian information of the target pedestrian, and counting the number of times of occurrence of each pedestrian which is the co-pedestrian with the target pedestrian in each co-pedestrian record;

and the client is used for displaying the records of the same pedestrians of which the occurrence times are greater than a second preset threshold value.

According to the method, the device, the server and the system for identifying the co-pedestrians, the images of the pedestrians acquired by the monitoring equipment are acquired, the images of the pedestrians are analyzed, the snapshot time and the snapshot place of each pedestrian in the images of the pedestrians are determined, and if the distance between the snapshot places of at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, the co-pedestrian record between the at least two pedestrians is generated once. Through analyzing the pedestrian image, can learn in the pedestrian image and take a candid photograph time and the candid photograph place of each pedestrian of taking a candid photograph, the interval in the final definite candid photograph place is in predetermineeing the within range, and two at least pedestrians of time interval in first predetermined time interval of time of taking a candid photograph are with the pedestrian, generate pedestrian's record once, need not to discern the pedestrian, note pedestrian's record between the pedestrian locally, for the inquiry personnel provide the inquiry foundation, pedestrian's record of local record can all be exported, but only export specific pedestrian's record, thereby can satisfy increasingly complicated pedestrian analysis demand.

Drawings

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

Fig. 1 is a schematic flow chart illustrating a method for identifying a fellow pedestrian according to an embodiment of the present invention;

FIG. 2 is a flowchart illustrating a peer identification method according to another embodiment of the present invention;

FIG. 3 is a flowchart illustrating a method for identifying a fellow pedestrian according to yet another embodiment of the present invention;

FIG. 4 is a flowchart illustrating a method for identifying a fellow pedestrian according to yet another embodiment of the present invention;

FIG. 5 is a flowchart illustrating a method for identifying a fellow pedestrian according to yet another embodiment of the present invention;

FIG. 6 is a schematic diagram of a single-screen capturing same-line scene according to an embodiment of the present invention;

FIG. 7 is a schematic diagram of a monitoring device acquiring a multi-picture snapshot peer scene according to an embodiment of the present invention;

fig. 8 is a schematic diagram of a scene captured by multiple monitoring devices in the same line in multiple pictures according to the embodiment of the present invention;

FIG. 9 is a schematic diagram of a scene of a capturing peer for filtering duplicate peers according to an embodiment of the present invention;

FIG. 10 is a scene diagram illustrating a principle of matching with pedestrians in four scenarios according to an embodiment of the present invention;

FIG. 11 is a flowchart illustrating a method for identifying a fellow pedestrian according to yet another embodiment of the present invention;

fig. 12 is a schematic structural diagram of a peer recognition device according to an embodiment of the present invention;

FIG. 13 is a schematic structural diagram of a peer recognition device according to another embodiment of the present invention;

FIG. 14 is a block diagram of a server according to an embodiment of the present invention;

fig. 15 is a schematic structural diagram of a fellow pedestrian recognition system according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

In order to meet increasingly complex pedestrian analysis requirements, the embodiment of the invention provides a method, a device, a server and a system for identifying a co-pedestrian. Next, a method for identifying a fellow pedestrian according to an embodiment of the present invention will be described.

The execution main body of the method for identifying the co-workers provided by the embodiment of the invention can be a server with core processing capability, and the mode for realizing the method for identifying the co-workers provided by the embodiment of the invention can be at least one of software, hardware circuit and logic circuit arranged in the execution main body.

As shown in fig. 1, a method for identifying a fellow pedestrian according to an embodiment of the present invention may include the following steps.

And S101, acquiring a pedestrian image acquired by monitoring equipment.

Under the application scenes of city safety, traffic monitoring, building monitoring and the like, one or more monitoring devices are generally distributed for real-time monitoring, when pedestrians enter a monitoring area, images of the pedestrians can be collected, and the monitoring devices can be traditional video cameras and can also be pedestrian snapshot devices. For the pedestrian snapshot device, once a pedestrian passes through the pedestrian snapshot device, the pedestrian snapshot device can automatically identify a pedestrian target, snapshot the pedestrian, and upload the captured pedestrian image to a server for identifying the same pedestrian. The monitoring equipment can be erected in a region where people flow greatly, and when the monitoring equipment is erected, conditions such as angle, height and brightness need to be guaranteed so as to meet the requirement for capturing pedestrians.

And S102, analyzing the pedestrian image, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian image.

The server for identifying the same pedestrian provides intelligent analysis service, after acquiring a pedestrian image, the pedestrian image can be analyzed, the analysis process is to extract the snapshot time and the snapshot place of each pedestrian in the pedestrian image, each pedestrian can be distinguished by the pedestrian feature, the pedestrian feature refers to the attribute feature of a pedestrian target, such as facial feature, clothes color, height, body type and the like, the snapshot time refers to the time for snapshotting the pedestrian target, the snapshot address can be read from the image attribute generally, the snapshot address refers to the position of the pedestrian target when the pedestrian target is snapshotted, the snapshot address can be represented by longitude and latitude information or other erection position information of monitoring equipment generally, and the snapshot address can also be the actual longitude and latitude of the pedestrian target certainly. The snapshot time and the snapshot place of one pedestrian can also form one piece of snapshot data, one pedestrian image or a plurality of pedestrian images are analyzed, and a plurality of pieces of snapshot data can be obtained.

And S103, if the distance between the snapshot places of at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians.

The embodiment of the invention provides a co-pedestrian judgment rule, namely, different pedestrians pass through a certain area successively (at intervals of several seconds and different pictures) or simultaneously (at the same picture), the same pedestrian is regarded as a co-pedestrian, and a co-pedestrian record among the pedestrians is generated. That is, if the distance between the capturing locations of at least two pedestrians is within a preset range (e.g., the same building, four directions of an intersection, the same square, etc.) and the time interval of the capturing time of the at least two pedestrians is within a first preset time interval (i.e., several seconds apart or simultaneously), the at least two pedestrians are identified as the same pedestrian, and the same-pedestrian record between the at least two pedestrians is generated once.

By applying the embodiment of the invention, the pedestrian images acquired by the monitoring equipment are acquired, the pedestrian images are analyzed, the snapshot time and the snapshot place of each pedestrian in the pedestrian images are determined, and if the distance between the snapshot places of at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, the co-pedestrian record between the at least two pedestrians is generated once. Through analyzing the pedestrian image, can learn in the pedestrian image and take a candid photograph time and the candid photograph place of each pedestrian of taking a candid photograph, the interval in the final definite candid photograph place is in predetermineeing the within range, and two at least pedestrians of time interval in first predetermined time interval of time of taking a candid photograph are with the pedestrian, generate pedestrian's record once, need not to discern the pedestrian, note pedestrian's record between the pedestrian locally, for the inquiry personnel provide the inquiry foundation, pedestrian's record of local record can all be exported, but only export specific pedestrian's record, thereby can satisfy increasingly complicated pedestrian analysis demand. And, every time at least two pedestrians are recognized to be the same-pedestrian, the same-pedestrian record between the at least two pedestrians can be generated once, and a plurality of results of the same-pedestrian record can be generated for different pedestrians.

Optionally, S103 may be specifically implemented by the following steps:

identifying at least two pedestrians with a preset distance between snapshot places and a first preset time interval between snapshot times; counting the times of continuously generating the co-pedestrian records between the at least two pedestrians in a second preset time interval; and if the times are greater than or equal to the first preset threshold value, keeping a one-time same-person record generated in a second preset time interval.

When identifying the same-pedestrian, on the basis of the original rule of the same-pedestrian, the filtering condition is added: the same at least two pedestrians are identified as the same pedestrian for a plurality of times in a short time, and only one record of the same pedestrian is reserved, namely in this case, the records of the same pedestrian are only recorded as 1 record of the same pedestrian no matter the same pedestrian is identified for a plurality of times.

Based on the co-pedestrian rule, after the co-pedestrian rule matching is carried out on the pedestrians under different scenes, the co-pedestrian records of the targets of the pedestrians can be obtained, and according to the co-pedestrian rule: if the number of times two or more pedestrians are continuously identified as co-pedestrians within a second predetermined time interval is greater than or equal to the first predetermined threshold (e.g., 20 co-pedestrians are identified a day), then the pedestrians can be identified as co-pedestrians, and the co-pedestrian record is recorded only once. When the record of the co-pedestrian is kept, the record of the co-pedestrian at the last time in the second preset time interval can be kept, and certainly, the records of the co-pedestrian at other times can also be kept, which is not specifically limited here. The data volume of the co-pedestrian records needing to be stored is reduced, the storage resources are effectively saved, and the pedestrians with the co-pedestrian relationship appearing for many times are marked as one-time co-pedestrian record through multiple identification, so that the misjudgment of the co-pedestrians can be reduced.

In an implementation manner of the embodiment of the present invention, the step of acquiring the pedestrian image acquired by the monitoring device may specifically be: acquiring at least one pedestrian image acquired by monitoring equipment; or acquiring a plurality of pedestrian images acquired by a plurality of monitoring devices.

Correspondingly, the step of analyzing the pedestrian image and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian image can be specifically as follows: analyzing a pedestrian image acquired by monitoring equipment, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian image, wherein the snapshot time is a timestamp of the monitoring equipment for acquiring the pedestrian image, and the snapshot place is the installation position of the monitoring equipment; or analyzing a plurality of pedestrian images acquired by one monitoring device, and determining the snapshot time and the snapshot place of each pedestrian in each pedestrian image, wherein the snapshot time is a timestamp of the monitoring device for acquiring each pedestrian image, and the snapshot place is the installation position of the monitoring device; or analyzing a plurality of pedestrian images acquired by a plurality of monitoring devices, and determining the snapshot time and the snapshot place of each pedestrian in each pedestrian image, wherein the snapshot time is a timestamp for acquiring each pedestrian image by each monitoring device, and the snapshot place is the installation position of each monitoring device for acquiring each pedestrian image.

If the distance between the snapshot places of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a pedestrian-sharing record between the at least two pedestrians, specifically: generating a co-pedestrian record between at least two pedestrians once aiming at least two pedestrians in one pedestrian image acquired by one monitoring device; or if the time interval of the snapshot time of at least two pedestrians in the multiple pedestrian images acquired by one monitoring device is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians at one time; or if the distance between the snapshot places of at least two pedestrians in the multiple pedestrian images acquired by the multiple monitoring devices is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a co-pedestrian record between the at least two pedestrians once.

One monitoring device can monitor a designated area range, a plurality of monitoring devices can also monitor the designated area range at the same time, and whether the plurality of monitoring devices monitor the designated area range can be judged by utilizing the installation positions (such as the installed longitude and latitude) of the monitoring devices. Because of the scene needs, a plurality of monitoring devices may be arranged at the same point, and the monitoring devices have complementary capturing capabilities, or the installation positions of the monitoring devices have geographical longitude and latitude differences, but the monitoring blind areas can be made up through cooperative work.

Thus, there are actually three snap-shot same pedestrian scenes. The method comprises the steps that a first snapshot same-pedestrian scene is a single-picture snapshot same-pedestrian scene, one monitoring device monitors one monitoring area, a pedestrian image collected by the monitoring device at a certain moment is analyzed, a timestamp of the monitoring device collecting the pedestrian image is used as snapshot time, and an installation position of the monitoring device is used as a snapshot place, so that when at least two pedestrians are snapshot in the pedestrian image, due to the fact that the same snapshot time (namely the timestamp) and the same snapshot place (namely the installation position) exist, at least two pedestrians in the pedestrian image are marked as a same-pedestrian record.

The second snapshot same-pedestrian scene is a same-monitoring-device multi-picture snapshot same-pedestrian scene, one monitoring device monitors one monitoring area, multiple pedestrian images collected by the monitoring device in a certain time period are analyzed, timestamps of the monitoring device collected the pedestrian images are used as snapshot time of all pedestrians in the pedestrian images, and the installation position of the monitoring device is used as a snapshot place of all pedestrians in the pedestrian images, so that when at least two pedestrians are snapshot in all pedestrian images, due to the fact that the same snapshot place (namely the installation position) exists, only whether the time interval of the snapshot time (namely the timestamp) of at least two pedestrians is within a first preset time interval is needed to be judged, and if yes, a same-pedestrian record between the at least two pedestrians is generated once.

The third snap-shot same-pedestrian scene is a multi-picture snap-shot same-pedestrian scene of different monitoring devices, and in the practical application scene, a plurality of monitoring devices are arranged, if the monitoring devices collect images of a plurality of pedestrians in a certain time period, analyzing the pedestrian images, taking the time stamp of each pedestrian image collected by each monitoring device as the snapshot time of each pedestrian in each pedestrian image, taking the installation position of each monitoring device as the snapshot place of each pedestrian in each pedestrian image, when at least two pedestrians are captured in each pedestrian image, it is necessary to determine whether the distance between the capturing locations (i.e., the installation locations) of the at least two pedestrians is within a preset range, and whether the time interval of the snapshot times (i.e. the time stamps) of the at least two pedestrians is within a first preset time interval, and if so, generating a co-pedestrian record between the at least two pedestrians once.

Based on the embodiment shown in fig. 1, an embodiment of the present invention further provides a method for identifying a fellow pedestrian, as shown in fig. 2, the method may include the following steps.

S201, acquiring a pedestrian image acquired by monitoring equipment.

S202, analyzing the pedestrian images, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian images.

S201 and S202 in the embodiment shown in fig. 2 are the same as S101 and S102 in the embodiment shown in fig. 1, and are not described again here.

And S203, identifying the pedestrian characteristics of each pedestrian, and clustering the pedestrians on the basis of the pedestrian characteristics of the pedestrians.

The server for identifying the same pedestrian also provides a pedestrian comparison service for comparing pedestrian features, wherein the comparison process can be carried out between the pedestrian features of the pedestrians, or the comparison can be carried out between the obtained pedestrians and a known pedestrian feature model, the comparison aims at clustering the pedestrians, and the pedestrians of the same type are clustered into one class. After clustering, different types of pedestrians can be distinguished by marking the identity label of each pedestrian.

In an implementation manner of the embodiment of the present invention, a specific manner of clustering pedestrians may be as follows: comparing the pedestrian characteristics of the pedestrians to obtain the pedestrian characteristic similarity among the pedestrians; marking the same identity label for the pedestrian with the pedestrian feature similarity larger than or equal to the preset similarity threshold; and marking different identity labels for the pedestrians with the pedestrian feature similarity smaller than the preset similarity threshold.

In the embodiment of the present invention, the pedestrian features may be compared with each other to obtain the similarity of the pedestrian features between pedestrians, for example, the pedestrian features such as the color of the clothes, the length of the hair, and the height of the pedestrian are respectively compared, and the similarity of the pedestrian features between two pedestrians is compared, and the higher the similarity is, the higher the possibility that the two pedestrians are the same pedestrian is, so in the embodiment of the present invention, a preset similarity threshold is set.

The pedestrian targets can be clustered through the labeling of the identity labels, and pedestrians in the same category are classified into the same category through labeling the same identity labels for the pedestrians with the pedestrian feature similarity being larger than or equal to the preset similarity threshold value.

Certainly, in the comparison process of the pedestrian characteristics, the pedestrian characteristics of each pedestrian can be compared with the known pedestrian characteristic models respectively, the pedestrian characteristic models are models of known pedestrians established based on experience, and the identity of the pedestrian can be known more clearly through comparison.

And S204, if the distance between the snapshot places of at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians.

After the pedestrians are clustered, the pedestrians are more accurately divided, pedestrians belonging to the same class are not recorded as one-time same-pedestrian record, but pedestrians belonging to different classes are recorded as one-time same-pedestrian record.

Based on the embodiment shown in fig. 1, an embodiment of the present invention further provides a method for identifying a fellow pedestrian, as shown in fig. 3, the method may include the following steps.

S301, acquiring a pedestrian image acquired by the monitoring equipment.

And S302, analyzing the pedestrian image, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian image.

S301 and S302 in the embodiment shown in fig. 3 are the same as S101 and S102 in the embodiment shown in fig. 1, and are not described again here.

And S303, storing the snapshot time and the snapshot place of each pedestrian to a first database.

After the snapshot time and the snapshot place of each pedestrian are obtained through analysis, the snapshot time and the snapshot place of each pedestrian can be stored in the first database, so that the purpose of big data storage is achieved, and a big data basis is provided for follow-up judgment of the same pedestrian.

S304, extracting the snapshot time and the snapshot place of at least two pedestrians from the first database.

S305, if the distance between the snapshot places of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a co-pedestrian record between the at least two pedestrians.

When the co-pedestrian recognition is carried out, the snapshot time and the snapshot place of at least two pedestrians can be directly extracted from the first database, whether the distance between the extracted snapshot places of the at least two pedestrians is within a preset range or not and whether the time interval of the snapshot time of the pedestrians is within a first preset time interval or not are judged, and if yes, the pedestrians are marked as a co-pedestrian record.

And S306, storing the co-workers to a second database.

And storing the records of the co-pedestrians by using the second database, and if a co-pedestrian query request is received subsequently, quickly finding out the co-pedestrian records containing the target pedestrians to be queried from the second database according to the co-pedestrian query request for feedback, or directly feeding back the pedestrians with the co-pedestrian relationship, wherein the feedback information can be sorted according to the co-pedestrian times.

Based on the embodiment shown in fig. 1, an embodiment of the present invention further provides a method for identifying a fellow pedestrian, as shown in fig. 4, the method may include the following steps.

S401, acquiring a pedestrian image acquired by monitoring equipment.

S402, analyzing the pedestrian image, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian image.

And S403, if the distance between the snapshot places of at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians.

S401 to S403 in the embodiment shown in fig. 4 are the same as S101 to S103 in the embodiment shown in fig. 1, and are not described again here.

S404, acquiring a co-pedestrian query request, wherein the co-pedestrian query request comprises pedestrian information of a target pedestrian to be queried.

When an inquirer (such as city security personnel, building management personnel and the like) has a co-pedestrian inquiry demand, a co-pedestrian inquiry request is sent to a server through an inquiry client (generally, a computer), wherein the co-pedestrian inquiry request comprises pedestrian information of a target pedestrian to be inquired, such as information of the name, the characteristics and the like of the target pedestrian.

S405, searching all co-pedestrian records in which the target pedestrian appears according to the pedestrian information of the target pedestrian, and counting the number of times of appearance of each pedestrian, which is the co-pedestrian with the target pedestrian, in each co-pedestrian record.

After receiving the co-pedestrian inquiry request, the server searches all co-pedestrian records of the target pedestrian from the recorded co-pedestrian records based on the pedestrian information of the target pedestrian, wherein the co-pedestrian records comprise the pedestrian information of other pedestrians having a co-pedestrian relationship with the target pedestrian, and counts the occurrence times of the other pedestrians.

And S406, outputting the records of the same pedestrians of which the occurrence times are greater than a second preset threshold value.

Since the inquirer generally focuses on the pedestrians who are often related to the target pedestrian, and there is a close relationship between the pedestrians and the target pedestrian, when the feedback is given to the inquirer, the record of the co-pedestrian whose occurrence number is greater than a second preset threshold value can be fed back, where the second preset threshold value can be set according to experience.

In the embodiment of the invention, after the query request of the co-workers is acquired, the co-worker record of the target pedestrian is searched from the recorded co-worker records in a targeted manner, and the co-worker record of which the occurrence frequency is greater than a second preset threshold value is output based on the co-walking frequency of each pedestrian and the target pedestrian. The accurate query result recorded by the same pedestrian can be output to the query staff in a targeted manner according to the real-time requirements of the query staff. And one-time co-pedestrian records can be generated, and results of a plurality of times of co-pedestrian records can be generated for different pedestrians, so that the query requirement of the query personnel is met.

For convenience of understanding, the peer identification method provided by the embodiment of the invention is described in detail below with reference to specific examples.

Fig. 5 is a schematic flow chart of a peer identification method according to an embodiment of the present invention.

The monitoring equipment reports the pedestrian image to a server providing central service after collecting the pedestrian image, the intelligent analysis service in the server analyzes the pedestrian images, determines the snapshot time and the snapshot place of each pedestrian in the pedestrian images, the pedestrian application program calls the snapshot data (the snapshot time and the snapshot place of each captured pedestrian), the pedestrian features are extracted, the comparison service compares the pedestrian features, and returns the comparison result to the pedestrian application program, sends the snapshot data and the comparison result to a data warehouse (a first database) for storage, when identifying the peer, extracting historical snapshot data from the data warehouse, and calculating big data, wherein the specific calculation process is as described in the above method embodiment, which is not described herein again, and finally outputting the calculation result to the database (second database) for storage.

The identification process of the same-pedestrian is introduced from three snap-shot same-pedestrian principle scenes.

As shown in fig. 6, a scene schematic diagram of a single-picture capturing and pedestrian, when a person 1 and a person 2 almost simultaneously pass through a monitoring device 1 in a monitoring area 1, the person 1 and the person 2 exist in the picture, captured pedestrian images are uploaded to a server, an intelligent analysis service analyzes the pedestrian images to obtain pedestrian characteristics 1 and pedestrian characteristics 2, the two pedestrian characteristics are compared through a comparison service, capturing records (including capturing time and capturing place) of the person 1 at time t1 and capturing records of the person 2 at time t1 are obtained, the capturing records are stored in a data warehouse, and according to rules of the same pedestrian: different pedestrians pass through a certain area successively (at intervals of several seconds and with different pictures) or simultaneously (with the same picture), the pedestrians are regarded as the same pedestrian, the person 1 and the person 2 can be determined to be the same pedestrian, and a same pedestrian record is generated.

As shown in fig. 7, a scene schematic diagram of a monitoring device capturing multiple pictures of a same pedestrian is shown, when a person 1 and a person 2 successively pass through the monitoring device 1 in a monitoring area 1, the monitoring device 1 captures the pictures, the person 1 and the person 2 exist in the pictures, two captured images of the pedestrian are uploaded to a server, an intelligent analysis service analyzes the images of the pedestrian to obtain a pedestrian characteristic 1 and a pedestrian characteristic 2, the two pedestrian characteristics are compared through a comparison service, a capture record of the person 1 at time t1 and a capture record of the person 2 at time t2 are obtained, the capture records are stored in a data warehouse, and according to rules of the same pedestrian: different pedestrians pass through a certain area successively (at intervals of several seconds, with different pictures) or simultaneously (with the same picture), the pedestrians are regarded as the same pedestrian, and because the difference value between t1 and t2 is smaller than a first preset time interval, the pedestrians 1 and 2 can be determined to be the same pedestrian, and a same pedestrian record is generated.

As shown in fig. 8, a scene schematic diagram of a co-pedestrian captured in multiple pictures is acquired by multiple monitoring devices, when a person 1 and a person 2 are captured by the monitoring device 1 and the monitoring device 2 in the same monitoring area 1 at times t1 and t2, respectively, two captured images of the pedestrian are uploaded to a server, an intelligent analysis service analyzes the images of the pedestrian to obtain a pedestrian characteristic 1 and a pedestrian characteristic 2, the two pedestrian characteristics are compared by a comparison service to obtain a capture record of the person 1 at time t1 and a capture record of the person 2 at time t2, the capture records are stored in a data warehouse, and according to rules of the co-pedestrian: different pedestrians pass through a certain area successively (at intervals of several seconds, with different pictures) or simultaneously (with the same picture), the pedestrians are regarded as the same pedestrian, and because the difference value between t1 and t2 is smaller than a first preset time interval, the pedestrians 1 and 2 can be determined to be the same pedestrian, and a same pedestrian record is generated.

Taking the same camera snapshot as an example, the embodiment of filtering the repeated pedestrian records and improving the effective pedestrian rate is described in detail below, as shown in fig. 9, when a person 1 and a person 2 almost simultaneously pass through the monitoring device 1 in the monitoring area 1 and are snapshot by the monitoring device 1, the person 1 and the person 2 exist in the picture, the captured pedestrian image is uploaded to the server, the pedestrian image is analyzed by the intelligent analysis service to obtain the pedestrian feature 1 and the pedestrian feature 2, the two pedestrian features are compared by the comparison service to obtain the snapshot record of the person 1 at the time t1 and the snapshot record of the person 2 at the time t1, and the snapshot records are stored in the data warehouse.

Due to the reasons of waiting for traffic lights, having a rest in place and the like, the personnel 1 and the personnel 2 stay in front of the monitoring equipment for a long time, the monitoring equipment 1 in the monitoring area carries out snapshot again, the captured pedestrian images are uploaded to the server, the pedestrian images are analyzed by the intelligent analysis service to obtain the pedestrian characteristics 1 and the pedestrian characteristics 2, the two pedestrian characteristics are compared through the comparison service to obtain the snapshot record of the personnel 1 at the time t2 and the snapshot record of the personnel 2 at the time t3, and the snapshot records are stored in the data warehouse.

Since the time interval of either t2 or t1 or t3 or t1 is smaller than the second preset time interval, the same-pedestrian recognition of the same group of pedestrians is considered to occur within a short time, and is retained only once and recorded as one-pedestrian record.

Based on the above pedestrian target co-walking rule, after the co-walking rule matching is performed on the data in different scenes (for example, matching in 4 scenes as shown in fig. 10), the co-walking record of each pedestrian can be obtained, the co-walking record is filtered according to the co-walking calculation rule (two or more pedestrians can see the co-walking of each other for multiple times in a period of time (which can be day, week or month)), and each pedestrian reaching the specified co-walking frequency threshold (for example, five records in one day) is used as the co-pedestrian and fed back to the inquirer.

In another method for identifying a co-pedestrian provided by the embodiment of the present invention, the execution subject is a system composed of a server and a database, as shown in fig. 11, the method may include the following steps.

S1101, acquiring a co-pedestrian query request, wherein the co-pedestrian query request comprises a target pedestrian to be queried.

And S1102, searching all co-pedestrian records with the target pedestrian from the database, and counting the occurrence times of the co-pedestrians, which are co-pedestrians with the target pedestrian, in each co-pedestrian record, wherein the co-pedestrian records of at least two pedestrians, in which the interval of the snapshot places in the pedestrian image acquired by the monitoring equipment is within a preset range and the time interval of the snapshot time is within a first preset time interval, are stored in the database.

And S1103, outputting the same-row record with the occurrence frequency larger than a preset threshold value.

In the embodiment of the invention, after the co-pedestrian query request is acquired, the co-pedestrian record of the target pedestrian is searched from the database in a targeted manner, and the co-pedestrian record with the occurrence frequency larger than a second preset threshold value is output based on the co-pedestrian frequency of each pedestrian and the target pedestrian. The database provides big data storage, provides big data query basis under the same-pedestrian query scene, ensures the integrity of data, and can provide guarantee for the accuracy of query results.

Corresponding to the above method embodiment, an embodiment of the present invention provides a peer identification apparatus, as shown in fig. 12, the apparatus may include:

the acquiring module 1210 is used for acquiring a pedestrian image acquired by monitoring equipment;

the analysis module 1220 is configured to analyze the pedestrian image to obtain the snapshot time and the snapshot place of each pedestrian in the pedestrian image;

the identification recording module 1230 is configured to generate a co-pedestrian record between the at least two pedestrians once if the distance between the snapshot locations of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval.

Optionally, the apparatus may further include:

the identification recording module 1230 may be specifically configured to:

Optionally, the obtaining module 1210 may be specifically configured to:

the analysis module 1220 may be specifically configured to:

alternatively, the first and second electrodes may be,

the identification recording module 1230 may be specifically configured to:

alternatively, the first and second electrodes may be,

Optionally, the apparatus may further include:

the identification recording module 1230 may be specifically configured to:

and storing the co-current person record to a second database.

Optionally, the identification recording module 1230 may be specifically configured to:

Optionally, the obtaining module 1210 may further be configured to:

the apparatus may further include:

An embodiment of the present invention further provides a device for identifying a fellow pedestrian, as shown in fig. 13, the device may include:

an obtaining module 1310, configured to obtain a peer query request, where the peer query request includes a target pedestrian to be queried;

the searching module 1320 is configured to search all co-pedestrian records where the target pedestrian appears from the database, and count the number of occurrences of each pedestrian who is the same pedestrian as the target pedestrian in each co-pedestrian record, where the database stores co-pedestrian records between at least two pedestrians whose capturing locations in the pedestrian image acquired by the monitoring device are within a preset range and whose capturing time interval is within a first preset time interval;

the output module 1330 is configured to output the peer record with the occurrence number greater than the preset threshold.

By applying the embodiment of the invention, after the co-pedestrian query request is acquired, the co-pedestrian record of the target pedestrian is searched from the database in a targeted manner, and the co-pedestrian record with the occurrence frequency larger than the second preset threshold value is output based on the co-pedestrian frequency with the target pedestrian. The database provides big data storage, provides big data query basis under the same-pedestrian query scene, ensures the integrity of data, and can provide guarantee for the accuracy of query results.

Embodiments of the present invention also provide a server, as shown in fig. 14, comprising a processor 1401 and a memory 1402, wherein,

a memory 1402 for storing a computer program;

the processor 1401 is configured to implement any one of the methods for identifying a fellow pedestrian according to the embodiments of the present invention when executing the computer program stored in the memory 1402.

The Memory may include a RAM (Random Access Memory) or an NVM (Non-Volatile Memory), such as at least one disk Memory. Optionally, the memory may also be at least one memory device located remotely from the processor.

The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also a DSP (Digital Signal Processing), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other Programmable logic device, discrete Gate or transistor logic device, discrete hardware component.

In this embodiment, the processor can realize that: the method comprises the steps of acquiring pedestrian images acquired by monitoring equipment, analyzing the pedestrian images, determining the snapshot time and the snapshot place of each pedestrian in the pedestrian images, and generating a same-pedestrian record between at least two pedestrians once if the interval between the snapshot places of at least two pedestrians is within a preset range and the time interval between the snapshot times of the at least two pedestrians is within a first preset time interval. Through analyzing the pedestrian image, can learn in the pedestrian image and take a candid photograph time and the candid photograph place of each pedestrian of taking a candid photograph, the interval in the final definite candid photograph place is in predetermineeing the within range, and two at least pedestrians of time interval in first predetermined time interval of time of taking a candid photograph are with the pedestrian, generate pedestrian's record once, need not to discern the pedestrian, note pedestrian's record between the pedestrian locally, for the inquiry personnel provide the inquiry foundation, pedestrian's record of local record can all be exported, but only export specific pedestrian's record, thereby can satisfy increasingly complicated pedestrian analysis demand. And, every time at least two pedestrians are recognized to be the same-pedestrian, the same-pedestrian record between the at least two pedestrians can be generated once, and a plurality of results of the same-pedestrian record can be generated for different pedestrians.

In addition, an embodiment of the present invention provides a computer-readable storage medium, where a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the computer program implements any peer identification method provided in the embodiment of the present invention.

In this embodiment, the computer-readable storage medium stores a computer program that executes the method for identifying a fellow pedestrian provided by the embodiment of the present invention when executed, and thus can implement: the method comprises the steps of acquiring pedestrian images acquired by monitoring equipment, analyzing the pedestrian images, determining the snapshot time and the snapshot place of each pedestrian in the pedestrian images, and generating a same-pedestrian record between at least two pedestrians once if the interval between the snapshot places of at least two pedestrians is within a preset range and the time interval between the snapshot times of the at least two pedestrians is within a first preset time interval. Through analyzing the pedestrian image, can learn in the pedestrian image and take a candid photograph time and the candid photograph place of each pedestrian of taking a candid photograph, the interval in the final definite candid photograph place is in predetermineeing the within range, and two at least pedestrians of time interval in first predetermined time interval of time of taking a candid photograph are with the pedestrian, generate pedestrian's record once, need not to discern the pedestrian, note pedestrian's record between the pedestrian locally, for the inquiry personnel provide the inquiry foundation, pedestrian's record of local record can all be exported, but only export specific pedestrian's record, thereby can satisfy increasingly complicated pedestrian analysis demand. And, every time at least two pedestrians are recognized to be the same-pedestrian, the same-pedestrian record between the at least two pedestrians can be generated once, and a plurality of results of the same-pedestrian record can be generated for different pedestrians.

Corresponding to the above embodiments, the embodiment of the present invention provides a peer identification system, as shown in fig. 15, the system includes a plurality of monitoring devices 1510 and a server 1520;

a monitoring device 1510 for acquiring pedestrian images;

the server 1520, is used for obtaining the pedestrian's picture that the supervisory equipment gathers; analyzing the pedestrian images, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian images; and if the distance between the snapshot places of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a co-pedestrian record between the at least two pedestrians once.

Optionally, the system may further include a first database and a second database;

Optionally, the system further comprises a client;

By applying the embodiment of the invention, the server acquires the pedestrian images acquired by the monitoring equipment, analyzes the pedestrian images, determines the snapshot time and the snapshot place of each pedestrian in the pedestrian images, and generates the same-pedestrian record between at least two pedestrians once if the distance between the snapshot places of at least two pedestrians is within the preset range and the time interval of the snapshot time of at least two pedestrians is within the first preset time interval. Through analyzing the pedestrian image, can learn in the pedestrian image and take a candid photograph time and the candid photograph place of each pedestrian of taking a candid photograph, the interval in the final definite candid photograph place is in predetermineeing the within range, and two at least pedestrians of time interval in first predetermined time interval of time of taking a candid photograph are with the pedestrian, generate pedestrian's record once, need not to discern the pedestrian, note pedestrian's record between the pedestrian locally, for the inquiry personnel provide the inquiry foundation, pedestrian's record of local record can all be exported, but only export specific pedestrian's record, thereby can satisfy increasingly complicated pedestrian analysis demand. And, every time at least two pedestrians are recognized to be the same-pedestrian, the same-pedestrian record between the at least two pedestrians can be generated once, and a plurality of results of the same-pedestrian record can be generated for different pedestrians.

For the server, the computer-readable storage medium and the system embodiment, since the contents of the related method are substantially similar to those of the foregoing method embodiment, the description is relatively simple, and the relevant points can be referred to the partial description of the method embodiment.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the embodiments of the apparatus, the server, the computer-readable storage medium, and the peer identification system, since they are substantially similar to the embodiments of the method, the description is simple, and the relevant points can be referred to the partial description of the embodiments of the method.

The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for identifying a fellow pedestrian, the method comprising:

acquiring a pedestrian image acquired by monitoring equipment;

analyzing the pedestrian image, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian image;

and if the distance between the snapshot places of at least two pedestrians is within a preset range and the time interval of the snapshot time of at least two pedestrians is within a first preset time interval, generating a co-pedestrian record between at least two pedestrians once.

2. The method according to claim 1, wherein before generating the co-pedestrian record between the at least two pedestrians once if the distance between the snapshot locations of the at least two pedestrians is within a preset range and the time interval of the snapshot times of the at least two pedestrians is within a first preset time interval, the method further comprises:

identifying a pedestrian characteristic of the respective pedestrian;

clustering the pedestrians based on the pedestrian features of the pedestrians;

if the interval in the snapshot place of two at least pedestrians is in the preset scope, and the time interval of the snapshot time of two at least pedestrians is in first preset time interval, then once generate the same pedestrian record between two at least pedestrians, include:

aiming at least two pedestrians belonging to different classes, if the distance between the snapshot places of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians once.

3. The method of claim 1, wherein the acquiring the pedestrian image collected by the monitoring device comprises:

the analyzing the pedestrian image to determine the snapshot time and the snapshot place of each pedestrian in the pedestrian image comprises the following steps:

analyzing a pedestrian image acquired by a monitoring device, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian image, wherein the snapshot time is a timestamp of the monitoring device for acquiring the pedestrian image, and the snapshot place is the installation position of the monitoring device;

alternatively, the first and second electrodes may be,

analyzing a plurality of pedestrian images acquired by a plurality of monitoring devices, and determining the snapshot time and the snapshot place of each pedestrian in each pedestrian image, wherein the snapshot time is a time stamp of each monitoring device for acquiring each pedestrian image, and the snapshot place is the installation position of each monitoring device for acquiring each pedestrian image;

generating a co-pedestrian record between at least two pedestrians once for at least two pedestrians in one pedestrian image acquired by one monitoring device;

alternatively, the first and second electrodes may be,

if the time interval of the snapshot time of at least two pedestrians in the multiple pedestrian images acquired by one monitoring device is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians once;

alternatively, the first and second electrodes may be,

if the distance between the snapshot places of at least two pedestrians in the multiple pedestrian images acquired by the monitoring devices is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval, generating a same-pedestrian record between the at least two pedestrians once.

4. The method of claim 1, wherein after analyzing the pedestrian image and determining the snapshot time and the snapshot location of each pedestrian in the pedestrian image, the method further comprises:

and storing the co-walking person record to a second database.

5. The method according to claim 1, wherein if the distance between the snapshot locations of at least two pedestrians is within a preset range and the time interval of the snapshot times of the at least two pedestrians is within a first preset time interval, generating a co-pedestrian record between the at least two pedestrians once comprises:

and if the times are greater than or equal to a first preset threshold value, keeping a one-time same-person record generated in the second preset time interval.

6. The method according to any one of claims 1 to 5, wherein after generating a co-pedestrian record between the at least two pedestrians once if the distance between the snapshot locations of the at least two pedestrians is within a preset range and the time interval of the snapshot times of the at least two pedestrians is within a first preset time interval, the method further comprises:

searching all co-pedestrian records in which the target pedestrian appears according to the pedestrian information of the target pedestrian, and counting the number of times of occurrence of each pedestrian in each co-pedestrian record, which is the co-pedestrian with the target pedestrian;

7. A method for identifying a fellow pedestrian, the method comprising:

searching all co-pedestrian records in which the target pedestrian appears from a database, and counting the occurrence times of the co-pedestrians, which are co-pedestrians with the target pedestrian, in each co-pedestrian record, wherein the co-pedestrian records of at least two pedestrians, in which the interval of the snapshot places in the pedestrian image acquired by the monitoring equipment is within a preset range and the time interval of the snapshot time is within a first preset time interval, are stored in the database;

8. A pedestrian recognition apparatus, comprising:

and the identification recording module is used for generating a co-pedestrian record between the at least two pedestrians once if the distance between the snapshot places of the at least two pedestrians is within a preset range and the time interval of the snapshot time of the at least two pedestrians is within a first preset time interval.

9. The apparatus of claim 8, further comprising:

the clustering module is used for identifying the pedestrian characteristics of all the pedestrians; clustering the pedestrians based on the pedestrian features of the pedestrians;

the identification recording module is specifically configured to:

10. The apparatus of claim 8, wherein the obtaining module is specifically configured to:

the analysis module is specifically configured to:

alternatively, the first and second electrodes may be,

the identification recording module is specifically configured to:

alternatively, the first and second electrodes may be,

11. The apparatus of claim 8, further comprising:

the identification recording module is specifically configured to:

and storing the co-walking person record to a second database.

12. The apparatus of claim 8, wherein the identification recording module is specifically configured to:

13. The apparatus according to any one of claims 8-12, wherein the obtaining module is further configured to:

the device further comprises:

the searching module is used for searching all co-pedestrian records in which the target pedestrian appears according to the pedestrian information of the target pedestrian, and counting the occurrence times of all pedestrians, which are co-pedestrians with the target pedestrian, in each co-pedestrian record;

14. A pedestrian recognition apparatus, comprising:

the system comprises an acquisition module, a query module and a query module, wherein the acquisition module is used for acquiring a co-pedestrian query request which comprises a target pedestrian to be queried;

the searching module is used for searching all co-pedestrian records with the target pedestrian from a database, counting the occurrence frequency of each pedestrian with the target pedestrian in each co-pedestrian record, wherein the co-pedestrian records of at least two pedestrians, the interval of the snap-shot places in the pedestrian images acquired by the monitoring equipment is within a preset range, and the time interval of the snap-shot time is within a first preset time interval, are stored in the database;

and the output module is used for outputting the same-row record of which the occurrence times is greater than a preset threshold value.

15. A pedestrian identification system is characterized by comprising a plurality of monitoring devices and a server;

the monitoring equipment is used for acquiring pedestrian images;

the server is used for acquiring the pedestrian image acquired by the monitoring equipment; analyzing the pedestrian image, and determining the snapshot time and the snapshot place of each pedestrian in the pedestrian image; and if the distance between the snapshot places of at least two pedestrians is within a preset range and the time interval of the snapshot time of at least two pedestrians is within a first preset time interval, generating a co-pedestrian record between at least two pedestrians once.

16. The system of claim 15, further comprising a first database and a second database;

the second database is used for storing the co-pedestrian record generated by the server.

17. The system of claim 15, further comprising a client;

the server is further used for acquiring a co-pedestrian query request, wherein the co-pedestrian query request comprises pedestrian information of a target pedestrian to be queried; searching all co-pedestrian records in which the target pedestrian appears according to the pedestrian information of the target pedestrian, and counting the number of times of occurrence of each pedestrian in each co-pedestrian record, which is the co-pedestrian with the target pedestrian;