CN112651335A

CN112651335A - Method, system, equipment and storage medium for identifying same-pedestrian

Info

Publication number: CN112651335A
Application number: CN202011566236.3A
Authority: CN
Inventors: 张倩; 王洁冰
Original assignee: Shenzhen Jizhi Digital Technology Co Ltd
Current assignee: Shenzhen Jizhi Digital Technology Co Ltd
Priority date: 2020-12-25
Filing date: 2020-12-25
Publication date: 2021-04-13
Anticipated expiration: 2040-12-25
Also published as: CN112651335B

Abstract

The method comprises the steps of obtaining an identification of a target person, a data search time range and a peer identification precision, obtaining a first set corresponding to the identification of the target person and the peer identification precision, searching data in the first set, obtaining a second set containing snapshot data matched with the data search time range, narrowing the data search range, searching data in the second set, obtaining a third set containing the target snapshot data, and further narrowing the search range. And according to the identification precision of the co-workers, the first set is obtained, the identification precision of the co-workers is used as one factor of the identification of the co-workers, and the flexibility of the identification of the co-workers is improved.

Description

Method, system, equipment and storage medium for identifying same-pedestrian

Technical Field

The present application relates to the field of object detection, and more particularly, to a method, system, device and storage medium for identifying a peer.

Background

In some scenarios (e.g., security or airport scenarios), it is desirable to be able to identify whether different pedestrians have a peer relationship.

However, how to identify whether different pedestrians have a peer-to-peer relationship becomes a problem.

Disclosure of Invention

In view of the above, the present application provides a method, a system, a device and a storage medium for identifying a peer, which are used for identifying a person having a peer relationship with a target person.

In order to achieve the above object, the following solutions are proposed:

a peer identification method, comprising:

acquiring the identification task information of the same-pedestrian, wherein the identification task information of the same-pedestrian at least comprises the following steps: the identification of the target person, the data searching time range and the identification precision of the same person;

acquiring a first set corresponding to the identification of the target person and the identification precision of the peer, wherein the first set comprises at least one piece of snapshot data related to the target person;

performing data search in the first set to obtain a second set containing snapshot data matched with the data search time range;

acquiring the snapshot time of the target person from the snapshot data in the second set, and performing data search in the second set to obtain a third set containing the target snapshot data, wherein the difference between the snapshot time of the person in the target snapshot data and the snapshot time of the target person is within a set time range;

and determining the persons having the same-row relationship with the target person according to the snapshot data in the third set.

The determining, according to the snapshot data in the third set, the person who has a peer relationship with the target person includes:

and taking the personnel identification contained in the snapshot data in the third set and the corresponding personnel as the personnel having the same-row relationship with the target person.

counting the number of times each personnel identifier contained in the snapshot data in the third set appears in the third set;

and sorting the times of the plurality of personnel identifications appearing in the third set in a descending order to obtain a sorting result, and taking the personnel corresponding to the first n personnel identifications arranged in the sorting result as the personnel having the same row relation with the target person, wherein n is an integer larger than 0.

Before the obtaining of the task information of the peer identification, the method further comprises the following steps:

acquiring snapshot information acquired by an image recognition device, wherein the snapshot information comprises an equipment identifier of each equipment in a snapshot range, an area identifier of each area, an identifier of each person, an association relationship between each equipment and the person and an association relationship between each area and the person, the equipment identifiers of each equipment are different, the area identifiers of each area are different, and the identifiers of each person are different;

carrying out first type structural processing on the snapshot information acquired by the image recognition device to obtain a data set conforming to a zset format;

the data set includes: the key is a device set of device identification, the score of the device set is a first time stamp containing time information of a person when the person is captured by the device, and the member is data containing the identification of the person captured by the device and the first time stamp; and a process for the preparation of a coating,

the key is an area set of area identification, the score of the area set is a second timestamp containing time information when the person appears in the area, and the member is data containing the identification of the person appearing in the area and the second timestamp; and a process for the preparation of a coating,

a first person set with keys for person identification, wherein the score of the first person set is a third timestamp containing time information when the person is captured by the equipment, and the member is data containing the identification of the equipment associated with the person and the third timestamp; and a process for the preparation of a coating,

a second person set with keys for person identification, wherein the score of the second person set is a fourth timestamp containing time information when the person appears in the area, and the member is data containing the identification of the area associated with the person and the fourth timestamp;

storing the data set.

If the accuracy of the peer identification is low, the obtaining of the first set corresponding to the identification of the target person and the accuracy of the peer identification includes:

acquiring at least one second person set corresponding to the low precision;

searching a second person set matched with the identification of the target person in at least one second person set;

if the target person is found, taking a second person set matched with the identification of the target person as a first set;

the performing data search in the first set to obtain a second set including a time range matching the data search includes:

searching for members with scores within the data search time range in the first set;

if the member is found, the found member is used as a first target member, an area set with keys matched with the area identifiers in the first target member is obtained from at least one area set, and the area set with the keys matched with the area identifiers in the first target member is used as a second set;

the data search in the second set to obtain a third set containing target snapshot data includes:

searching members in the second set, wherein the difference between the time of the person appearing in the area and the time of the target person appearing in the area is within a set time range;

and if the member is found, taking the found member as a second target member, and taking a set containing the second target member as a third set.

If the accuracy of the identification of the co-workers is high, the obtaining of the first set corresponding to the identification of the target person and the accuracy of the identification of the co-workers comprises:

acquiring at least one first person set corresponding to the high precision;

searching a first person set matched with the identification of the target person in at least one first person set;

if the target person is found, taking a first person set matched with the identification of the target person as a first set;

if the device identification is found, the found member is used as a first target member, a device set with keys matched with the device identification in the first target member is obtained from at least one device set, and the device set with the keys matched with the device identification in the first target member is used as a second set;

searching for members in the second set, wherein the difference between the time when the person is captured by the equipment and the time when the target person is captured by the equipment is within a set time range;

The method further comprises the following steps:

and carrying out second type structural processing on the snapshot information acquired by the image recognition device to obtain a data set conforming to a character string format, wherein the data set conforming to the character string format comprises at least one piece of data in the character string format, and a key in the data in the character string format is an identifier of a person and has a value of a storage address of a snapshot image containing the person.

The method further comprises the following steps:

searching data in a character string format matched with the identification of the target person in the data set conforming to the character string format;

if the target character is found, acquiring a snapshot image containing the target character according to a storage address in the data in the character string format matched with the identification of the target character, and outputting the snapshot image containing the target character;

and/or searching data in a character string format matched with the identification of the co-pedestrian of the target character in the data set conforming to the character string format;

if the target character is found, acquiring a snapshot image of the peer including the target character according to a storage address in the data in the character string format matched with the identifier of the peer of the target character, and outputting the snapshot image of the peer including the target character.

A peer identification system, comprising:

the first acquisition module is used for acquiring the peer identification task information, and the peer identification task information at least comprises: the identification of the target person, the data searching time range and the identification precision of the same person;

a second obtaining module, configured to obtain a first set corresponding to the identifier of the target person and the recognition accuracy of the peer person, where the first set includes at least one piece of snapshot data related to the target person;

the first searching module is used for searching data in the first set to obtain a second set containing snapshot data matched with the data searching time range;

the second searching module is used for acquiring the snapshot time of the target person from the snapshot data in the second set, searching data in the second set to obtain a third set containing the target snapshot data, and the difference between the snapshot time of the person in the target snapshot data and the snapshot time of the target person is within a set time range;

and the determining module is used for determining the persons having the same-row relationship with the target person according to the snapshot data in the third set.

The determining module is specifically configured to:

The system further comprises:

the third acquisition module is used for acquiring snapshot information acquired by the image recognition device, wherein the snapshot information comprises an equipment identifier of each equipment in a snapshot range, an area identifier of each area, an identifier of each person, an association relationship between each equipment and the person and an association relationship between each area and the person, the equipment identifiers of each equipment are different, the area identifiers of each area are different, and the identifiers of each person are different;

the first processing module is used for carrying out first type structured processing on the snapshot information acquired by the image recognition device to obtain a data set conforming to a zset format;

and the storage module is used for storing the data set.

The second obtaining module is specifically configured to, if the accuracy of identifying the peer is low accuracy, obtain at least one second people set corresponding to the low accuracy;

the first search module is specifically configured to:

the second search module is specifically configured to:

The second obtaining module is specifically configured to:

if the identification precision of the co-workers is high precision, at least one first person set corresponding to the high precision is obtained;

the first search module is specifically configured to:

the second search module is specifically configured to:

The system further comprises:

and the second processing module is used for carrying out second type structural processing on the snapshot information acquired by the image recognition device to obtain a data set conforming to the character string format, wherein the data set conforming to the character string format comprises at least one piece of data in the character string format, and the key in the data in the character string format is the identifier of a person and the value is the storage address of the snapshot image containing the person.

The system further comprises:

the third searching module is used for searching the data in the character string format matched with the identification of the target person in the data set conforming to the character string format;

the third acquisition module is used for acquiring a snapshot image containing the target character according to a storage address in the data in the character string format matched with the identifier of the target character if the data in the character string format matched with the identifier of the target character is found;

the first output module is used for outputting a snapshot image containing the target person;

and/or the presence of a gas in the gas,

the fourth searching module is used for searching the data in the character string format matched with the identification of the co-pedestrian of the target character in the data set conforming to the character string format;

the fourth acquisition module is used for acquiring the snap-shot image of the peer containing the target character according to the storage address in the data in the character string format matched with the identifier of the peer of the target character;

and the second output module is used for outputting the snapshot image of the same pedestrian containing the target character.

A computer device comprising a processor and a memory, said memory having stored therein at least one instruction, at least one program, a set of codes, or a set of instructions, which is loaded and executed by said processor to implement a peer identification method as claimed in any preceding claim.

A computer-readable storage medium, having stored thereon a computer program which, when executed by a processor, implements a method of peer identification as in any one of the above.

According to the technical scheme, the identification of the target person, the data search time range and the identification precision of the co-workers are obtained by obtaining the identification task information of the co-workers, the first set corresponding to the identification of the target person and the identification precision of the co-workers is obtained, data search is carried out in the first set, the second set containing snapshot data matched with the data search time range is obtained, the data search range is narrowed, data search is carried out in the second set, the third set containing the target snapshot data is obtained, and the search range is further narrowed.

And according to the identification precision of the co-workers, the first set is obtained, the identification precision of the co-workers is used as one factor of the identification of the co-workers, and the flexibility of the identification of the co-workers is improved.

Drawings

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

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

fig. 2 is a schematic flowchart of a method for identifying a fellow pedestrian according to embodiment 2 of the present application;

fig. 3 is a schematic flowchart of a method for identifying a fellow pedestrian according to embodiment 3 of the present application;

fig. 4 is a schematic flowchart of a method for identifying a fellow pedestrian according to embodiment 4 of the present application;

fig. 5 is a schematic flowchart of a method for identifying a fellow pedestrian according to embodiment 5 of the present application;

fig. 6 is a schematic logical structure diagram of a peer identification system according to an embodiment of the present disclosure.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, 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 application.

In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.

Referring to fig. 1, a flowchart of a method for identifying a fellow pedestrian provided in embodiment 1 of the present application, the method may be applied to an electronic device, the present application does not limit a product type of the electronic device, and as shown in fig. 1, the method may include, but is not limited to, the following steps:

step S11, obtaining the identification task information of the same-person, wherein the identification task information of the same-person at least comprises the following steps: the identification of the target person, the data search time range and the identification precision of the co-workers.

In this embodiment, a front-end interface may be set, which supports a user to input peer identification task information, such as an identifier of a target task, a data search time range, and peer identification precision, in the front-end interface. Accordingly, the peer recognition system can actively acquire the peer recognition task information from the front-end interface.

Of course, the user may also upload the query parameters through an interface provided by the peer recognition system, and the peer recognition system analyzes the query parameters to obtain the peer recognition task information.

The query parameter may or may not include the identification accuracy of the same person. Under the condition that the inquiry parameters include the identification precision of the co-pedestrian, the inquiry parameters are analyzed, and the identification precision of the co-pedestrian can be directly obtained. In the case that the query parameter does not include the peer identification accuracy, the default peer identification accuracy (e.g., low accuracy) may be used as the peer identification accuracy of the peer identification task of this time.

Step S12, obtaining a first set corresponding to the identification of the target person and the accuracy of the peer recognition, where the first set includes at least one piece of snapshot data related to the target person.

It will be appreciated that data sets corresponding to different pedestrian recognition accuracies differ from one another.

Since the peer recognition task is to recognize the person having the peer relationship with the target person, it is necessary to acquire snapshot data corresponding to the identification of the target task. When the task of identifying the fellow passenger has a requirement for accuracy of identifying the fellow passenger, a first set corresponding to the identification of the target person and the accuracy of identifying the fellow passenger is acquired. Specifically, the snapshot data matching the identification of the target person may be acquired from the image recognition apparatus. And capturing data corresponding to the recognition accuracy of the same pedestrian from the capturing data matched with the identification of the target person, and taking a set formed by the acquired data as a first set.

It is understood that the image recognition device takes a snapshot of the person within the snapshot range and recognizes the snapshot information from the snapshot image. The snapshot information may include, but is not limited to: the equipment identification of each equipment, the area identification of each area and the identification of each personnel in the snapshot range, wherein the equipment identification of each equipment is different, the area identification of each area is different, and the identification of each personnel is different.

Wherein the first set includes at least one piece of snapshot data related to the target person. The snapshot data associated with the target person may include, but is not limited to: data associated with the area in which the target person appears (e.g., the area identification and the time at which the target person was captured in the area), and data associated with the device that captured the target person (e.g., the device identification and the time at which the device captured the target person).

And step S13, performing data search in the first set to obtain a second set containing snapshot data matched with the data search time range.

Performing data search in the first set to obtain a second set including snapshot data matching the data search time range, which may be understood as:

finding snapshot data in the first set that matches the data finding time range;

and if the snapshot data is found, taking a set containing the snapshot data matched with the data search time range as a second set.

Step S14, determining the snapshot time of the target person according to the snapshot data in the second set, and performing data search in the second set to obtain a third set containing target snapshot data, wherein the difference between the snapshot time of the person in the target snapshot data and the snapshot time of the target person is within a set time range.

The second set is obtained by searching data in the first set, and the data in the first set is snapshot data related to the target person, so that the data in the second set is snapshot data related to the target person and matched with the data searching time range. Since the snapshot data related to the target person includes the snapshot time of the target person, the snapshot time of the target person can be acquired from the snapshot data in the second set.

In this embodiment, the set time range may be set as needed, and is not limited in this application. Wherein the smaller the setting of the time range is, the more accurate the acquired snapshot data in the third set is.

And step S15, determining the persons having the same-row relationship with the target person according to the snapshot data in the third set.

Because the snapshot data contains information such as the identification of the person, the person who has the peer relationship with the target person can be determined according to the fear data in the third set.

In the method, the identification of the target person, the data search time range and the identification precision of the co-workers are obtained by obtaining the identification task information of the co-workers, the first set corresponding to the identification of the target person and the identification precision of the co-workers is obtained, data search is carried out in the first set, the second set containing snapshot data matched with the data search time range is obtained, the data search range is narrowed, data search is carried out in the second set, the third set containing the target snapshot data is obtained, and the search range is further narrowed.

And moreover, the inquiry parameters can be analyzed to obtain the identification precision of the co-workers, and the first set is obtained according to the identification precision of the co-workers, so that the identification precision of the co-workers is taken as one of the factors for identifying the co-workers, and the flexibility of identifying the co-workers is improved.

As another alternative embodiment of the present application, referring to fig. 2, a flowchart of an embodiment 2 of a peer identification method provided by the present application is provided, where this embodiment mainly relates to a refinement scheme of the peer identification method described in the above embodiment 1, as shown in fig. 2, the method may include, but is not limited to, the following steps:

step S21, obtaining the identification task information of the same-person, wherein the identification task information of the same-person at least comprises the following steps: the identification of the target person, the data search time range and the identification precision of the co-workers.

Step S22, obtaining a first set corresponding to the identification of the target person and the accuracy of the peer recognition, where the first set includes at least one piece of snapshot data related to the target person.

And step S23, performing data search in the first set to obtain a second set containing snapshot data matched with the data search time range.

Step S24, capturing the capturing time of the target person from the capturing data in the second set, and performing data search in the second set to obtain a third set containing the target capturing data, wherein the difference between the capturing time of the person in the target capturing data and the capturing time of the target person is within a set time range.

The detailed procedures of steps S21-S24 can be found in the related descriptions of steps S11-S14 in embodiment 1, and are not repeated herein.

And step S25, taking the personnel identification contained in the snapshot data in the third set and corresponding personnel as the personnel having the same-row relationship with the target person.

In this embodiment, the person identifier included in the snapshot data in the third set and the corresponding person are used as the person having the peer relationship with the target person, so that omission of the person having the peer relationship with the target person can be avoided, and reliability of peer identification is ensured.

As another alternative embodiment of the present application, referring to fig. 3, a flowchart of an embodiment 3 of a peer identification method provided by the present application is provided, where this embodiment mainly relates to a refinement scheme of the peer identification method described in the foregoing embodiment 1, as shown in fig. 3, the method may include, but is not limited to, the following steps:

step S31, obtaining the identification task information of the same-person, wherein the identification task information of the same-person at least comprises the following steps: the identification of the target person, the data search time range and the identification precision of the co-workers.

Step S32, obtaining a first set corresponding to the identification of the target person and the accuracy of the peer recognition, where the first set includes at least one piece of snapshot data related to the target person.

And step S33, performing data search in the first set to obtain a second set containing snapshot data matched with the data search time range.

Step S34, capturing the capturing time of the target person from the capturing data in the second set, and performing data search in the second set to obtain a third set containing the target capturing data, wherein the difference between the capturing time of the person in the target capturing data and the capturing time of the target person is within a set time range.

The detailed procedures of steps S31-S34 can be found in the related descriptions of steps S11-S14 in embodiment 1, and are not repeated herein.

And step S35, counting the occurrence times of each personnel identifier contained in the snapshot data in the third set.

And step S36, performing descending order sorting on the times of the plurality of personnel identifications appearing in the third set to obtain a sorting result, and taking the personnel corresponding to the first n personnel identifications arranged in the sorting result as the personnel having the same-row relationship with the target person, wherein n is an integer larger than 0.

In this embodiment, the size of n may be set as needed, and is not limited in this application.

In this embodiment, the number of times of occurrence in the third set is counted by counting each person identifier included in the snapshot data in the third set, and the number of times of occurrence in the third set of the plurality of person identifiers is sorted in a descending order to obtain a sorting result, and persons corresponding to the top n person identifiers arranged in the sorting result are used as persons having a peer relationship with the target person, so that the person having a high occurrence frequency is used as a person having a peer relationship with the target person, and the accuracy of identifying the peer person is improved.

As another alternative embodiment of the present application, referring to fig. 4, a flowchart of an embodiment 4 of a peer identification method provided by the present application is provided, where this embodiment is mainly an extension of the peer identification method described in the foregoing embodiment 1, and as shown in fig. 4, the method may include, but is not limited to, the following steps:

and step S41, capturing the snapshot information collected by the image recognition device.

In this embodiment, the image recognition device can capture the person within the capture range, and recognize the capture information from the captured image. The snapshot information may include, but is not limited to: the method comprises the steps of capturing equipment identification of each equipment in a range, area identification of each area, identification of each person, association relation between each equipment and the person and association relation between each area and the person, wherein the equipment identification of each equipment is different, the area identification of each area is different, and the identification of each person is different.

The association relationship between each device and the person can be understood as: information of the person captured by each device.

The association relationship between each region and the person can be understood as: information of the persons present in each area.

And step S42, carrying out first type structural processing on the snapshot information acquired by the image recognition device to obtain a data set conforming to the zset format.

the key is a second person set of person identifications, the score of the second person set is a fourth time stamp containing time information when the person appears in the area, and the member is data containing the identification of the area associated with the person and the fourth time stamp.

In this embodiment, the form of the member is not limited. For example, the members of the set of devices may be represented as: first timestamp @ identification of person captured by the device; the members in the set of regions may be represented as: a second timestamp @ identification of the person present in the area; the members of the first set of people may be represented as: third timestamp @ identification of the device associated with the person; the members of the second set of people may be represented as: fourth timestamp @ identification of the area associated with the person.

Next, an example of an apparatus set with keys as apparatus identifiers, an area set with keys as area identifiers, a first person set with keys as person identifiers, and a second person set with keys as person identifiers is described:

for example, the captured snapshot information collected by the image recognition device includes: fear information of two persons is involved. The snapshot information includes: the identification of two personnel, the identification of 5 equipment in the snapshot scope, the identification of 3 regions in the snapshot scope, the time stamp that contains the time information when every personnel was snapshot by equipment to and contain the time stamp of the time information that every personnel appears in the region, the incidence relation of every equipment and personnel and the incidence relation of every region and personnel. The identities of the two people are denoted personid1 and personid2, respectively; the identities of the 5 devices are denoted as camera 1, camera 2, camera 3, camera 4, camera 5, respectively; the identities of the 3 regions are denoted area 1, area 2, area 3, respectively.

The data in the device set (which may be expressed as: key: CID:1) with key camera 1 can be seen in table 1,

TABLE 1

The data in the set of regions with the key camera 1 (which can be expressed as: key: AID:1), can be seen in table 2,

TABLE 2

Score of	Member
		20200101120000	20200101120000@personid1
20200101120001	20200101120001@personid1
		20200101121001	20200101120001@personid1
20200101121101	20200101121101@personid1
		20200101121201	20200101121201@personid1
20200101120000	20200101120000@personid2
		20200101120001	20200101120001@personid2
20200101121001	20200101120001@personid2
		20200101121101	20200101120001@personid2
20200101121201	20200101120001@personid2

The data in the first person set (which may be expressed as: keys: PID:1: C) with keys of personid1 can be seen in table 3,

TABLE 3

Score of	Member
		20200101120000	20200101120000@cameraid1
20200101120001	20200101120001@cameraid1
		20200101121001	20200101120001@cameraid2
20200101121101	20200101121101@cameraid3
		20200101121201	20200101121201@cameraid3

The data in the second human set (which may be expressed as keys: PID:1: a) with a key of personid1 can be seen in table 4,

TABLE 4

Score of	Member
		20200101120000	20200101120000@areaid1
20200101120001	20200101120001@areaid1
		20200101121001	20200101120001@areaid2
20200101121101	20200101121101@areaid2
		20200101121201	20200101121201@areaid2

It should be noted that, the members in the device set are data including the identifier of the person captured by the device and the first timestamp, the members in the area set are data including the identifier of the person appearing in the area and the second timestamp, the members in the first person set are data including the identifier of the device associated with the person and the third timestamp, and the members in the second person set are data including the identifier of the area associated with the person and the fourth timestamp, which can ensure that the members in each set do not have duplication, reduce the workload of data search, improve the efficiency of data query, and further improve the efficiency of identifying the same person.

And step S43, storing the data set.

Step S44, obtaining the identification task information of the same-person, wherein the identification task information of the same-person at least comprises the following steps: the identification of the target person, the data search time range and the identification precision of the co-workers.

The detailed process of step S44 can be referred to the related description of step S11 in embodiment 1, and is not repeated here.

And step S45, if the accuracy of identifying the fellow pedestrian is low, acquiring at least one second people set corresponding to the low accuracy.

If the recognition accuracy of the same person is low, the same person can be located to the region level for searching, and therefore at least one second person set corresponding to the low accuracy can be obtained.

Step S46, searching for a second person set matching the identification of the target person from at least one second person set.

If the search result is found, step S47 is executed.

And step S47, taking the second person set matched with the identification of the target person as a first set.

Steps S45-S47 are a specific implementation of step S12 in example 1.

Step S48, in the first set, searching for members whose scores are within the data search time range.

If the search result is found, step S49 is executed.

Step S49, taking the searched member as a first target member, obtaining an area set with keys matched with the area identifiers in the first target member from at least one area set, and taking the area set with the keys matched with the area identifiers in the first target member as a second set.

Steps S48-S49 are a specific implementation of step S13 in example 1.

And step S410, searching members in the second set, wherein the difference between the time of the person appearing in the area and the time of the target person appearing in the area is within a set time range.

If the search result is found, step S411 is executed.

Step S411, the searched member is used as a second target member, and the set containing the second target member is used as a third set.

Step S411 is a specific implementation manner of step S14 in example 1.

Referring now to tables 1, 2, 3 and 4, steps S44-S411 are described as an example, for example, if the identification of the target person is personId1, the data search time is in the range of 2020-01-0112:00:00 to 2020-01-0113: 00:00, and the accuracy of the peer identification is low, then the key is found according to the personId (with a value of personId1) of the target person: PID:1: A. According to the data search time range 2020-01-0112:00:00 to 2020-01-0113: 00:00, at a key: PID 1: A, taking members with the scores of [20200101120000,20200101130000], and obtaining a member set and an area set; according to the value in the area set, finding AID:1 and AID: 2 two sets of regions. And traversing AID:1 and AID: 2 members in two area sets, find the time information and key contained in the time stamp: PID 1: a member of 5 seconds of difference between the time information contained in the time stamps corresponding to personid 1. For example, if there is a value 20200101120000@ area 1 in the member set, it indicates that the person corresponding to personid1 appears in area 1 at the time indicated by 20200101120000, and then a member whose difference from the time indicated by 20200101120000 is not more than 5 seconds is found, and the person recorded in the found member identifies the corresponding person as a candidate co-pedestrian.

And S412, determining the persons having the same-row relationship with the target person according to the snapshot data in the third set.

The detailed process of step S412 can be referred to the related description of step S15 in embodiment 1, and is not described herein again.

In this embodiment, after the snapshot information acquired by the image recognition device is acquired, a second type of structural processing may be performed on the snapshot information acquired by the image recognition device to obtain a data set conforming to a character string format, where the data set conforming to the character string format includes at least one piece of data in the character string format, and a key in the data in the character string format is an identifier of a person and a value of the key is a storage address of a snapshot image including the person.

Correspondingly, after acquiring the information of acquiring the peer identification task, the following steps can be further executed:

In this embodiment, the acquisition of the snapshot image of the pedestrian including the target person, the acquisition of the snapshot image of the target person, and the output of the snapshot image are supported, so that more processing can be realized according to the snapshot image.

In this embodiment, if the accuracy of the peer identification is low, the data can be located to the region level for data search, so that the data coverage is wide, omission of people having a peer relationship with the target person is avoided, and the reliability of the peer identification is ensured.

As another alternative embodiment of the present application, referring to fig. 5, a flowchart of an embodiment 5 of a peer identification method provided by the present application is provided, where this embodiment is mainly an extension of the peer identification method described in the foregoing embodiment 1, and as shown in fig. 5, the method may include, but is not limited to, the following steps:

and step S51, capturing the snapshot information collected by the image recognition device.

And step S52, carrying out first type structural processing on the snapshot information acquired by the image recognition device to obtain a data set conforming to the zset format.

And step S53, storing the data set.

Step S54, obtaining the identification task information of the same-person, wherein the identification task information of the same-person at least comprises the following steps: the identification of the target person, the data search time range and the identification precision of the co-workers.

The detailed procedures of steps S51-S54 can be referred to the related descriptions of steps S41-S44 in embodiment 4, and are not described herein again.

And step S55, if the identification precision of the co-workers is high, acquiring at least one first person set corresponding to the high precision.

If the recognition accuracy of the same person is high, the same person can be located to the equipment level for searching, and at least one first person set corresponding to the high accuracy can be obtained.

Step S56, finding a first person set matching the identification of the target person from at least one first person set.

If the search result is found, step S57 is executed.

Step S57, the first person set matching the identification of the target person is used as the first set.

Steps S55-S57 are a specific implementation of step S12 in example 1.

Step S58, in the first set, searching for members whose scores are within the data search time range.

If the search result is found, step S59 is executed.

Step S59, taking the searched member as a first target member, obtaining an equipment set with a key matched with the equipment identifier in the first target member from at least one of the equipment sets, and taking the equipment set with the key matched with the equipment identifier in the first target member as a second set.

Steps S58-S59 are a specific implementation of step S13 in example 1.

Step 510, searching members in the second set, wherein the difference between the time when the person is captured by the device and the time when the target person is captured by the device is within a set time range.

If the search result is found, step S511 is executed.

Step S511, the searched member is used as a second target member, and the set including the second target member is used as a third set.

Step S511 is a specific implementation manner of step S14 in example 1.

Referring now to tables 1, 2, 3 and 4, steps S54-S511 are described as an example, for example, if the identification of the target person is personId1, the data search time range is 2020-01-0112:00:00 to 2020-01-0113: 00:00, and the accuracy of the identification of the peer is high, then the key is found according to the personId (with a value of personId1) of the target person: PID:1: C. According to the data search time range 2020-01-0112:00:00 to 2020-01-0113: 00:00, at a key: taking members with the fraction of [20200101120000,20200101130000] from PID 1: C to obtain a member set and a camera set; according to the values in the camera set, finding CID:1 and CID: 2 two device sets. And traversing CID 1 and CID: 2 members of two device sets, find the time information and key contained in the time stamp: PID 1: members in which the difference between the time information contained in the time stamps corresponding to personid1 is within 5 seconds. For example, if there is a value 20200101120000@ camera 1 in the member set, it indicates that the person corresponding to personid1 is captured by the device corresponding to camera 1 at the time represented by 20200101120000, and then a member whose difference from the time represented by 20200101120000 is not more than 5 seconds is found, and the person corresponding to the person identifier recorded in the found member is identified as a candidate co-pedestrian.

And S512, determining the persons having the same-row relationship with the target person according to the snapshot data in the third set.

The detailed process of step S512 can be referred to the related description of step S15 in embodiment 1, and is not described herein again.

In this embodiment, if the accuracy of the co-worker identification is high, the co-worker identification can be located to the equipment level for data search, so that the accuracy of the data is ensured, and the accuracy of the co-worker identification is ensured.

Next, the system for identifying a fellow pedestrian provided by the present application will be described, and the system for identifying a fellow pedestrian described below and the method for identifying a fellow pedestrian described above may be referred to in correspondence with each other.

Referring to fig. 6, the peer recognition system includes: a first obtaining module 100, a second obtaining module 200, a first lookup module 300, a second lookup module 400, and a determining module 500.

A first obtaining module 100, configured to obtain peer identification task information, where the peer identification task information at least includes: the identification of the target person, the data search time range and the identification precision of the co-workers.

A second obtaining module 200, configured to obtain a first set corresponding to the identifier of the target person and the recognition accuracy of the peer person, where the first set includes at least one piece of snapshot data related to the target person.

The first searching module 300 is configured to perform data searching in the first set, and obtain a second set including snapshot data that matches the data searching time range.

The second searching module 400 is configured to obtain the snapshot time of the target person from the snapshot data in the second set, perform data search in the second set, and obtain a third set including target snapshot data, where a difference between the snapshot time of the person in the target snapshot data and the snapshot time of the target person is within a set time range.

And the determining module 500 is configured to determine, according to the snapshot data in the third set, a person who has a peer relationship with the target person.

In this embodiment, the determining module 500500 may be specifically configured to:

In this embodiment, the system may further include:

and the storage module is used for storing the data set.

In this embodiment, the second obtaining module 200 may be specifically configured to obtain at least one second person set corresponding to the low accuracy if the identification accuracy of the peer is low accuracy;

the first searching module 300 may be specifically configured to:

the second searching module 400 may be specifically configured to:

In this embodiment, the second obtaining module 200 may be specifically configured to:

the first searching module 300 may be specifically configured to:

the second searching module 400 is specifically configured to:

In this embodiment, the system may further include:

and/or the presence of a gas in the gas,

In another embodiment of the present application, there is provided a computer device comprising a processor and a memory, said memory having stored therein at least one instruction, at least one program, a set of codes, or a set of instructions, said at least one instruction, said at least one program, said set of codes, or said set of instructions being loaded and executed by said processor to implement the peer identification method as described in any one of method embodiments 1-5 above.

In another embodiment of the present application, a computer-readable storage medium is provided, on which a computer program is stored, which computer program, when being executed by a processor, carries out the peer identification method as described in any one of the above method embodiments 1-5.

Finally, it should also be 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.

The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

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

2. The method of claim 1, wherein determining people who have a peer relationship with the target person based on the snapshot data in the third set comprises:

3. The method of claim 1, wherein determining people who have a peer relationship with the target person based on the snapshot data in the third set comprises:

4. The method of claim 1, wherein prior to obtaining peer identification task information, further comprising:

storing the data set.

5. The method of claim 4, wherein obtaining the first set corresponding to the identity of the target person and the peer recognition accuracy if the peer recognition accuracy is low comprises:

acquiring at least one second person set corresponding to the low precision;

6. The method of claim 4, wherein obtaining the first set corresponding to the identity of the target person and the peer recognition accuracy if the peer recognition accuracy is high comprises:

acquiring at least one first person set corresponding to the high precision;

7. The method of claim 4, further comprising:

8. The method of claim 7, further comprising:

9. A pedestrian recognition system, comprising:

10. A computer device comprising a processor and a memory, the memory having stored therein at least one instruction, at least one program, a set of codes, or a set of instructions, the at least one instruction, the at least one program, the set of codes, or the set of instructions being loaded and executed by the processor to implement a peer identification method as claimed in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method for identifying a fellow person according to any one of claims 1 to 8.