CN110852372B

CN110852372B - Data association method, device and equipment and readable storage medium

Info

Publication number: CN110852372B
Application number: CN201911080827.7A
Authority: CN
Inventors: 熊一鸣; 杨森
Original assignee: Beijing Aibee Technology Co Ltd
Current assignee: Beijing Aibee Technology Co Ltd
Priority date: 2019-11-07
Filing date: 2019-11-07
Publication date: 2022-05-31
Anticipated expiration: 2039-11-07
Also published as: CN110852372A

Abstract

The application provides a data association method, a data association device, equipment and a readable storage medium, WiFi data and face data are obtained, the WiFi data comprise an MAC address of the equipment connected with a WiFi network, and the face data comprise face images. The WiFi data is divided into at least two parts, and the WiFi data of the target part is divided into a plurality of partitions. And obtaining matched data groups by determining face data matched with the WiFi data in each partition, wherein any one matched data group comprises a matched MAC address and a face. Further, the score of the target matching data set is obtained by an operation result of the first numerical value (i.e., the total number of times the target matching data set appears in all parts) and the second numerical value (i.e., the total number of parts including the target matching data set). Since the first numerical value and the second numerical value are respectively in direct proportion to the matching degree of the MAC address and the face in the target matching data group, the face matched with the MAC address is obtained according to the score of each matching data group.

Description

Data association method, device and equipment and readable storage medium

Technical Field

The present application relates to the field of information processing, and more particularly, to a data association method, apparatus, device, and readable storage medium.

Background

At present, some valuable information can be obtained through big data analysis, wherein the association of various data is one of the big data analysis, so that the valuable information can be obtained through the association of various data, and a basis can be provided for the decision of a user and the like.

Disclosure of Invention

The application provides a data association method and a data association device, and aims to determine an association relationship between WiFi data and face data.

In order to achieve the above object, the present application provides the following technical solutions:

a method of data association, comprising:

the method comprises the steps of obtaining WiFi data and face data, wherein the WiFi data comprise MAC addresses of equipment connected with a WiFi network, and the face data comprise face images;

dividing the WiFi data into at least two parts, wherein the WiFi data of the at least two parts are not intersected and have the same data quantity, and the WiFi data of the devices on the same floor are included in the same part;

dividing WiFi data of a target part into a plurality of partitions according to a plurality of preset signal strength thresholds, wherein the target part is any one part;

obtaining matched data groups by determining face data matched with the WiFi data in each partition, wherein any one matched data group comprises a matched MAC address and a face;

counting a target matching data set in the target part, wherein the target matching data set is any one of the matching data sets, and the total times of occurrence of the target matching data set in the partition of the target part are counted;

taking an operation result of a first numerical value and a second numerical value as a score of the target matching data set, wherein the first numerical value is the total number of times that the target matching data set appears in all the parts, and the second numerical value is the total number of all the parts including the target matching data set;

and obtaining the face matched with the MAC address according to the scores of the matched data groups.

Optionally, dividing WiFi data of the target portion into a plurality of partitions according to a plurality of preset signal strength thresholds includes:

acquiring a preset signal intensity threshold sequence, wherein the signal intensity threshold sequence comprises N signal intensity thresholds from large to small;

and dividing WiFi data larger than a target signal intensity threshold value into one partition to obtain N partitions.

Optionally, the signal strength threshold is determined according to the distribution of the signal strength of the device connected to the WiFi network, and/or the distance between the device and the WiFi access point.

Optionally, taking the operation result of the first numerical value and the second numerical value as the score of the target matching data set includes:

and taking the product of the first numerical value and the second numerical value as the score of the target matching data set.

Optionally, determining face data matched with the WiFi data in each partition to obtain a matching data set, including:

using WiFi data with the same MAC address in the WiFi data as a group to obtain a WiFi data group;

calculating matching scores between the MAC address contained in each WiFi data group and each face in the face image respectively, wherein the matching scores are positively correlated with the degree that equipment to which the MAC address belongs and the degree that the face to which the face image belongs respectively indicate the same person;

and respectively taking the faces corresponding to the preset number of matching scores which are sorted from large to small according to the matching scores from the target MAC address and the matching scores among the faces as the faces matched with the target MAC address, wherein the target MAC address is any one MAC address.

Optionally, determining a matching score between a MAC address included in any WiFi data group and any face in the face image includes:

determining face data meeting preset conditions with each WiFi data in the WiFi data group from the face data of the face, and using the face data as a face data group associated with the WiFi data; for any piece of WiFi data, the preset conditions include: the time belongs to a preset time range; the preset time range is a time range formed by preset time lengths before and after the generation time of the WiFi data is taken as a time midpoint;

respectively calculating the matching score between each WiFi data and the associated face data group to obtain the matching score of each WiFi data;

and determining the matching score between the MAC address contained in the WiFi data group and the face according to the matching score of each piece of WiFi data.

Optionally, calculating a matching score between any piece of WiFi data and the associated face data group includes:

calculating the matching score of the WiFi data and each piece of face data in the associated face data group; the matching score of the WiFi data and any piece of face data in the associated face data group is in negative correlation with the gap; the difference is a difference between the first distance and the second distance; the first distance is the distance between the position in the WiFi data and the router to which the WiFi data belongs; the second distance is the distance between the position in the piece of face data and the router to which the piece of WiFi data belongs;

and adding the WiFi data and the matching score of each piece of face data in the associated face data group to obtain a value which is used as the matching score between the WiFi data and the associated face data group.

Optionally, after the calculating the matching score between each piece of WiFi data and the associated face data group respectively to obtain the matching score of each piece of WiFi data, and before determining the matching score between the MAC address included in the WiFi data group and the face according to the matching score of each piece of WiFi data, the method further includes:

determining WiFi data of overlapping time periods in the WiFi data belonging to different routers as a group of WiFi data to be processed from the WiFi data group;

sequencing the moments of the WiFi data to be processed according to a preset sequence to obtain sequenced WiFi data to be processed;

determining the weight of the matching score of each piece of WiFi data in the sorted WiFi data to be processed; the weight of the matching score of any piece of WiFi data is negatively related to the target distance; the target distance is the distance between the router to which the WiFi data and the adjacent WiFi data belong respectively;

respectively weighting the matching scores of each group of WiFi data to be processed, wherein the value obtained by weighting and summing the matching scores of any group of WiFi data to be processed is the matching score of the group of WiFi data to be processed;

adding the matching scores of the WiFi data except the WiFi data to be processed of each group in the WiFi data group and the matching scores corresponding to the WiFi data to be processed of each group respectively;

the determining the matching score between the MAC address contained in the WiFi data group and the face according to the matching score of each piece of WiFi data comprises the following steps:

and taking the added value as the matching score between the MAC address contained in the WiFi data group and the face.

A data association apparatus, comprising:

the data acquisition module is used for acquiring WiFi data and face data, the WiFi data comprises an MAC address of equipment connected with a WiFi network, and the face data comprises a face image;

the division module is used for dividing the WiFi data into at least two parts, wherein the WiFi data of the at least two parts are not intersected and have the same data volume, and the WiFi data of the equipment on the same floor are included in the same part;

the device comprises a partitioning module, a processing module and a processing module, wherein the partitioning module is used for dividing WiFi data of a target part into a plurality of partitions according to a plurality of preset signal strength thresholds, and the target part is any one part;

the matching module is used for obtaining matching data groups by determining face data matched with the WiFi data in each partition, and any one matching data group comprises a matched MAC address and a face;

a counting module for counting a total number of times that a target matching data group in the target portion appears in the partition of the target portion, the target matching data group being any one of the matching data groups;

a scoring module, configured to use an operation result of a first numerical value and a second numerical value as a score of the target matching data set, where the first numerical value is a total number of times that the target matching data set appears in all the portions, and the second numerical value is a total number of all the portions including the target matching data set;

and the association module is used for obtaining the face matched with the MAC address according to the scores of the matched data groups.

A data association device, comprising: a memory and a processor;

the memory is used for storing programs;

the processor is configured to execute the program to implement the steps of the data association method.

A readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the data association method as described above.

According to the technical scheme, the data association method, the data association device, the data association equipment and the readable storage medium divide the acquired WiFi data into at least two parts according to the position, and divide any part, namely the target part, into N partitions according to the signal intensity. And for any partition, matching the acquired face data with the WiFi data, and determining the face data matched with the WiFi data in each partition to obtain a matched data group, wherein the MAC address and the face contained in the matched data group appearing in any partition are the MAC address and the face with high matching degree in the partition. It is apparent that for any one of the matching data sets, i.e., the target matching data set, the number of occurrences in the partition (i.e., the first value) is proportional to the degree to which the MAC address matches the face in the target matching data set. And, the total number (i.e. the second value) of all the parts including the target matching data set is also in proportion to the matching degree of the MAC address and the face in the target matching data set. Therefore, the operation result of the first numerical value and the second numerical value is used as the score of the target matching data set. It will be appreciated that the score of the target match data set is proportional to the degree to which the MAC address matches the face in the target match data set. Therefore, the obtained face matched with the MAC address has higher accuracy according to the scores of all the matched data groups. In addition, the method partially divides WiFi data, data communication among different positions (such as floors) is achieved, matching results can be verified mutually, and reliability of the matching results is guaranteed.

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 view of an application scenario of a data association method according to an embodiment of the present application;

fig. 2 is a flowchart of a data association method according to an embodiment of the present application;

fig. 3 is a flowchart of a method for determining a matched MAC address and a face according to an embodiment of the present application;

fig. 4 is a schematic structural diagram of a data association apparatus according to an embodiment of the present application;

fig. 5 is a schematic structural diagram of a data association device according to an embodiment of the present application.

Detailed Description

Fig. 1 is a schematic view of an application scenario of a data association method provided in an embodiment of the present application, including: the data association device 103 comprises a first device 101, a second device 102 and the data association device 103, wherein the first device 101 is used for providing WiFi data, the second device 102 is used for providing face data, and the data association device 103 is used for determining that the WiFi data provided by the first device 101 and the face data provided by the second device 102 are associated to match the WiFi data and the face data belonging to the same person.

In this embodiment of the application, the WiFi data provided by the first device and the face data provided by the second device refer to: the WiFi data and the face data are generated in a preset space in a specified place, wherein the preset space can be divided into floors. For example, the designated place is the XX mall, the preset spaces in the designated place can be one layer, two layers, three layers, and four layers of the mall, respectively, and the WiFi data provided by the first device and the face data provided by the second device refer to WiFi data and face data generated by the one layer, the two layers, the three layers, and the four layers in the XX mall.

It should be noted that fig. 1 is only one optional application scenario of the embodiment of the present application, and the embodiment of the present application may also be applied to other scenarios, which is not limited to this embodiment of the present application.

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.

Fig. 2 is a schematic flow chart of a data association method provided in an embodiment of the present application, and as shown in fig. 2, the method may specifically include:

s201, WiFi data and face data are obtained.

Specifically, the WiFi data is data that the router in the appointed place is scanned to the equipment under each moment, the router produced, wherein, the WiFi data specifically includes: the MAC address of the device connected to the WiFi network, the time at which the router generates the WiFi data, the signal strength, and the location. The device may be a mobile phone, the embodiment does not limit the specific form of the device, and the position indicates a position of the device relative to the scanned router. For example, a device with MAC address a scans a router B in the second layer of the mall, and the distance between the device with MAC address a and the router B can be determined according to the signal strength scanned by the device. The designated place can be set according to the actual business scene, for example, set as the designated market.

Optionally, WiFi data may be acquired by a WiFi probe. In this embodiment, WiFi data of one MAC address at one time is referred to as one piece of WiFi data.

The face data is positions of all people in a specified place at all times, and specifically, the face data includes: the method comprises the steps of obtaining a face image, shooting time of the face image and the position of the face in a specified place when the face in the face image is shot.

Specifically, the manner of acquiring the face data may include: the method comprises the steps of identifying face images and shooting time of the face images from video streams shot by cameras installed in a specified place, carrying out three-dimensional reconstruction according to images in the video streams, determining positions of people contained in the identified face images at the time in the specified place, and forming face data by the identified face images, the shooting time of the face images and the positions of the people contained in the face images at the shooting time in the specified place.

For convenience of description, in this embodiment, face data constituted by the position of one face at one time is referred to as one piece of face data.

Further, the position in the obtained WiFi data and the position in the face data are unified to a preset reference coordinate system to obtain unified WiFi data and unified face data. The implementation process of the position unification is the prior art, and is not described herein again.

S202, the WiFi data is divided into at least two parts.

The WiFi data of the devices located on the same floor are included in the same portion, and the WiFi data of each portion do not intersect, that is, each WiFi data value is included in only one portion. Moreover, the data amount of the WiFi data included in each divided part is equivalent, that is, the data amount included in different parts cannot be too different. In this regard, a threshold of the data amount difference may be preset, and the WiFi data may be divided according to the total amount of the WiFi data and the data amount of the WiFi data of each floor.

Taking the division of two parts as an example, assuming that the WiFi data acquired in S202 includes WiFi data of one, two, three, and four layers in the XX mall, the step may divide the WiFi data into two parts based on the position of each piece of WiFi data, that is, the first part includes all WiFi data of the two layers, and the second part includes all WiFi data of the one, three, and four layers. Wherein the amount of data in the first portion is comparable to the amount of data in the second portion.

S203, dividing WiFi data of the target part into a plurality of partitions according to a plurality of preset signal strength thresholds.

In this step, the process of dividing any one part of WiFi data into a plurality of partitions is the same, and for convenience of description, a method for dividing WiFi data of a target part into a plurality of partitions is introduced with any one part as the target part, which specifically includes:

first, a preset signal strength threshold sequence is obtained.

The signal strength threshold sequence comprises N signal strength thresholds from large to small. Each piece of WiFi data of the target portion includes a signal strength and a location, so the signal strength threshold may be determined according to a distribution of signal strengths of devices connected to the WiFi network and/or a distance between the device and the WiFi access point. For example, when the signal strength threshold is determined according to the signal strength distribution of the devices connected to the WiFi network, and the signal strength distribution of the WiFi data of the target portion is between-75 dB and-45 dB, the threshold may be determined to be-80 dB, -70dB, -60dB, and-50 dB, respectively. Thus, the preset signal strength threshold sequence is determined to be-80 dB, -70dB, -60dB, -50 dB.

Further, the WiFi data larger than the target signal strength threshold value are divided into one partition, and N partitions are obtained. The target signal intensity threshold is any signal intensity threshold of the signal intensity threshold sequence, and N is the number of the signal intensity thresholds. For example, the signal intensity threshold sequence is { -80dB, -70dB, -60dB, -50dB }, WiFi data with the signal intensity greater than-80 dB is divided into a subarea, WiFi data with the signal intensity greater than-70 dB is divided into a subarea, WiFi data with the signal intensity greater than-60 dB is divided into a subarea, and WiFi data with the signal intensity greater than-50 dB is divided into a subarea.

Based on the above method, any portion may be divided into N partitions based on a signal strength threshold sequence.

And S204, obtaining a matched data group by determining the face data matched with the WiFi data in each partition.

Specifically, each partition includes several pieces of WiFi data, each piece of WiFi data includes a MAC address of a device connected to the WiFi network, and it can be understood that the MAC addresses of the same device (e.g., a handset) when connected to the WiFi network are the same. Each piece of face data includes a face image, and each face image at least indicates one face, that is, the same face may be included in a plurality of face images. Therefore, by determining the degree of association between each WiFi data and each face data, the degree of matching between each MAC address and a face can be further determined. Namely, the association degree between the WiFi data containing the MAC address and the face data containing the face is high, and the matching degree between the MAC address and the face is high.

In the step, the association relationship between the WiFi data and the face data in each partition is determined, the face matched with any MAC address is determined, and the matched MAC address and the face form a matched data group, so that the matched data group of each partition is obtained.

In this embodiment, the method for determining the matched MAC address and face may include: the method comprises the steps of taking WiFi data of the same MAC address in WiFi data as a group to obtain WiFi data groups, calculating matching scores between the MAC address contained in each WiFi data group and each face in a face image, wherein the matching scores are positively correlated with the degree that equipment to which the MAC address belongs and the face to which the face image belongs respectively indicate the same person, and regarding each MAC address, the faces respectively corresponding to preset number of matching scores in the sequence from large to small of the matching scores in the matching scores between the MAC address and each face are taken as the faces correlated with the MAC address.

S205, counting the total times of the target matching data group in the target part appearing in the partition of the target part.

The target part is any one part, and the target matching data set is any one matching data set. It will be appreciated that different partitions may derive the same matching data set, for example, by determining face data for which WiFi data in a first partition matches, a matching data set Z. The matching data set Z is also obtained by determining face data matched with the WiFi data in the second partition. This step counts the target match data set (e.g., Z), the total number of occurrences in the target portion, i.e., the number of partitions in which the target portion appears in the target match data set.

S206, taking the operation result of the first numerical value and the second numerical value as the score of the target matching data set.

The first value is the total number of times that the target matching data set appears in all parts, and the second value is the total number of all parts including the target matching data set.

For example, WiFi data is divided into a first portion and a second portion, where each portion is divided into a first partition, a second partition, a third partition, and a fourth partition. For a target matched data set Z, the first value is the sum of the number of occurrences of the target matched data set Z in the first portion and the number of occurrences in the second portion. The second value is 1 if the target match data set Z is present only in the first portion or only in the second portion, and the second value is 2 if the target match data set Z is present in both the first portion and the second portion.

Obviously, the larger the first value is, the more times the MAC address and the face are determined to be matched, that is, the first value is proportional to the degree of matching between the MAC address and the face in the target matching data. And, the larger the second value is, the more positions the MAC address and the face are determined to be matched, that is, the second value is also proportional to the degree of matching between the MAC address and the face in the target matching data. Therefore, a product of the first value and the second value may be calculated as a fraction of the target matching data set. From which the scores of the respective matching data sets can be derived.

And S207, obtaining the face matched with the MAC address according to the scores of the matched data groups.

It will be appreciated that the score of any matching data set is proportional to the degree to which the matching data set includes a MAC address and a face. Therefore, according to the scores of the matching data groups, the matching degree of the MAC address and the face can be judged, and the MAC address and the face with high matching degree can be obtained.

This step can be performed in a number of ways, two alternative ways, a first way and a second way, are described below.

In the first mode, the matching data sets with the scores smaller than the preset score threshold are eliminated, the matching data sets with the scores larger than or equal to the preset score threshold are used as final matching data sets, and the MAC addresses and the faces in each final matching data set are determined to be matched MAC addresses and faces.

And in the second mode, the matching data groups comprising the same MAC address are sequenced according to the scores, the front preset matching data group is taken as a final matching data group, and the MAC address and the face in each final matching data group are determined as the matched MAC address and the face.

According to the technical scheme, the data association method, the data association device, the data association equipment and the readable storage medium divide the acquired WiFi data into at least two parts according to the position, and divide any part, namely the target part, into N partitions according to the signal intensity. And for any partition, matching the acquired face data with the WiFi data, and determining the face data matched with the WiFi data in each partition to obtain a matched data group, wherein the MAC address and the face contained in the matched data group appearing in any partition are the MAC address and the face with high matching degree in the partition. It is apparent that, for any matching data set, i.e., the target matching data set, the number of occurrences in the partition (i.e., the first value) is proportional to the degree to which the MAC address matches the face in the target matching data set. And, the total number (i.e. the second value) of all the parts including the target matching data set is also in proportion to the matching degree of the MAC address and the face in the target matching data set. Therefore, the operation result of the first numerical value and the second numerical value is used as the score of the target matching data set. It will be appreciated that the score of the target match data set is proportional to the degree to which the MAC address matches the face in the target match data set. Therefore, the obtained face matched with the MAC address has higher accuracy according to the scores of all the matched data groups. In addition, the method partially divides WiFi data, data communication among different positions (such as floors) is achieved, matching results can be verified mutually, and reliability of the matching results is guaranteed.

Further, the method divides the WiFi data of each part into partitions, and WiFi data in different partitions intersect, that is, WiFi data is repeatedly included in multiple partitions, and is usually included in only one partition due to low signal strength of noise data, so that the weight of WiFi data is increased and the weight of noise data is reduced. Based on this, interference of the noise data is reduced.

Further, because a large amount of data interference exists in the obtained WiFi data or the face data, in the embodiment of the present application, the unified WiFi data and the unified face data in S201 are preprocessed, so that the preprocessed WiFi data and the preprocessed face data are obtained.

Alternatively, the pre-processing may comprise two aspects, a first aspect and a second aspect respectively.

Wherein, the first aspect is: and removing data irrelevant to data association and data with poor quality from the unified WiFi data to obtain the removed WiFi data. And removing data irrelevant to data association from the unified face data to obtain the removed face data. The second aspect is: and respectively carrying out data sampling on the removed face data and the removed WiFi data to obtain preprocessed WiFi data and preprocessed face data.

For WiFi data: the processing of the first aspect on the unified WiFi data includes: removing data not related to the data association from the unified WiFi data includes: and removing WiFi data with duration not within a preset duration range from the unified WiFi data. Removing poor quality data from the consolidated WiFi data includes: and removing the WiFi data with the signal intensity smaller than a preset intensity threshold value from the unified WiFi data.

The preset duration range is composed of a duration upper limit value and a duration lower limit value, wherein the duration upper limit value is the duration for distinguishing designated personnel from non-designated personnel. Taking a specific place as a mall and a customer as a person (i.e., a specific person) associated with the data as an example, since the time lengths of the customer and the non-customer (staff) in the mall are different, a time length value for distinguishing the staff from the customer may be determined as the upper limit value of the time length, for example, the upper limit value of the time length may be 8 hours. Specifically, the specific value of the duration upper limit needs to be set according to an actual situation, and the specific value of the duration upper limit is not limited in this embodiment.

The lower limit value of the duration is used for distinguishing whether the personnel indicated by the equipment to which the MAC address belongs in the WiFi data is the personnel in the specified place. Also taking a given place as a mall and a customer as a given person as an example, since the device of a person who passes by the mall and does not enter the mall may also scan a router in the mall, however, the person who passes by the mall and does not enter the mall is not a person (non-given person) relevant to data association. Moreover, the duration of the WiFi data of the person who passes through the mall and does not enter the mall is shorter than that of the person in the mall, so a time length value for distinguishing whether the person indicated by the device to which the MAC address belongs in the WiFi data is a person in the mall may be set as the lower limit value of the time length. For example, the lower limit of the time period is 2 minutes.

The WiFi data with the signal strength smaller than the threshold represents WiFi data with poor signal strength, specifically, a specific value of the threshold needs to be set according to an actual situation, for example, the value may be-90 dB, and the specific value of the threshold is not limited in this embodiment.

Aiming at the face data: in this step, the processing of the first aspect on the unified face data includes: and removing the face data except the face data of the pre-designated personnel from the face data, wherein the designated personnel are personnel required to perform data association in the actual service scene. Taking a designated place as an example, the customer is a designated person.

Specifically, the process of removing the face data other than the face data of the designated person from the unified face data includes: and recognizing the face data except the face data of the appointed person from the unified face data, and deleting the recognized face data. The process of identifying the face data other than the face data of the designated person from the unified face data may include: the face information of the appointed person is counted in advance, the counted face information is identified from the unified face data, and then the face data except the face data of the appointed person in the unified face data is obtained. Of course, this embodiment only provides a way of recognizing the face data other than the face data of the designated person from the unified face data, and in practice, other ways may also be used, and this embodiment does not limit the specific recognition way.

After the removed WiFi data and the removed face data are obtained, the removed WiFi data and the removed face data are respectively processed in the second aspect, that is, data sampling is respectively performed, so that the preprocessed face data and the preprocessed WiFi data are obtained.

In this embodiment, in order to reduce the computing resources and improve the computing efficiency, the processing of the second aspect is performed on the removed face data, specifically, the removed face data is sampled. The specific sampling process comprises the following steps: and taking the initial time of the removed face data, the time of which the time length with the initial time is integral multiple of the first preset time length, and the ending time as the face sampling time. And deleting the face data except the face data at the face sampling moment in the removed face data to obtain the preprocessed face data. The specific value of the first preset duration may be set according to an actual service scenario, and the value of the first preset duration is not limited in this embodiment.

And the processing of the second aspect on the removed WiFi data is to perform data sampling on the removed WiFi data. Specifically, the process of performing data sampling on the removed WiFi data includes: and taking the initial time, the time with the time length between the initial time and the end time of the removed WiFi data as integral multiple of a second preset time length and the WiFi sampling time, and deleting the WiFi data except the WiFi data at the WiFi sampling time in the removed WiFi data to obtain the preprocessed WiFi data. The specific value of the second preset duration needs to be set according to a specific service scenario, and the specific value of the second preset duration is not limited in this embodiment.

In practice, there is a certain delay between the moment of generating the WiFi data and the moment of scanning (the moment of scanning the router generating the WiFi data), in this embodiment, a delay duration is set based on the delay, for example, the delay duration is 5s, of course, in practice, the preset delay duration may also be other values, and this embodiment does not limit the specific value of the preset delay duration. In this embodiment, when the first preset duration is the same as the second preset duration and is equal to the preset delay duration, the subsequent steps are performed based on the preprocessed WiFi data and the preprocessed face data, and the accuracy of the correlation result between the obtained WiFi data and the face data is improved.

The preprocessing method can be seen that the WiFi data and the face data are respectively processed in the first aspect, so that the data quality of the removed WiFi data and the removed face data is improved for data association, and the accuracy of the finally obtained face result associated with each MAC address in the WiFi data is ensured to a certain extent. Meanwhile, after the removed WiFi data and the removed face data are processed in the second aspect, redundant data in the removed WiFi data and the removed face data can be greatly reduced, the preprocessed WiFi data and the preprocessed face data are obtained, a process of obtaining a final correlation result based on the preprocessed WiFi data and the preprocessed face data in a subsequent calculation mode can be achieved, more calculation resources can be saved, and calculation efficiency can be improved.

Fig. 3 is a method for determining a matched MAC address and a face according to an embodiment of the present application, where the method for determining face data matched with WiFi data of any partition in any part is the same, and the method for determining a matched MAC address and a face in S204 is further described by taking any partition as an example in the embodiment of the present application, which may specifically include:

s301, the WiFi data with the same MAC address in the WiFi data are used as a group, and a WiFi data group is obtained.

The WiFi data is WiFi data in any partition, and the WiFi data may be preprocessed WiFi data or unified WiFi data.

S302, calculating matching scores between the MAC address contained in each WiFi data group and each face in the face image.

It can be understood that the size of the matching score is positively correlated with the degree to which the device to which the MAC address belongs and the face to which the face image belongs indicate the same person, respectively. That is, the larger the matching score of the MAC address with the face in the face image, the greater the degree to which the MAC address and the face indicate the same person.

In this step, the process of determining the matching score between the MAC address included in each WiFi data group and each face in the face data is the same, and for convenience of description, the process of determining the matching score between the MAC address included in the WiFi data group and each face is described by taking any one WiFi data group and any one face as an example. The method specifically comprises the following steps A1-A5:

and A1, determining a face data group matched with each WiFi data in the WiFi data group.

In this step, the process of determining the face data group associated with each piece of WiFi data is the same, and for convenience of description, the determination of the face data group associated with the piece of WiFi data from the face data of the face is described by taking any piece of WiFi data as an example.

Specifically, the face data meeting the preset condition is determined from the face data of the face to be the face data group associated with the WiFi, wherein the preset condition includes: the generation time belongs to a preset time range. The preset time range comprises an upper limit value and a lower limit value, wherein the upper limit value is the time before the generation time of the WiFi data is the time midpoint of the preset time length, and the lower limit value is the time after the generation time of the WiFi data is the time midpoint of the preset time length.

And aiming at any piece of WiFi data, determining that a face data group meeting preset conditions in the face data of the face is a face data group associated with the WiFi data through the step. In this embodiment, it needs to be determined whether the device to which the MAC address included in the piece of WiFi data belongs and the face included in the determined face data group respectively indicate the same person. In practice, sometimes the device to which the MAC address of the piece of WiFi data belongs and the face image included in the associated face data set indicate that the face is not obviously indicative of the same person, if it is determined that the match score is meaningless according to the following steps. For example, a piece of WiFi data and face data meeting a preset condition generated by a router on the third floor of the mall are a set of face data on the first floor of the mall.

Therefore, in this step, in order to improve the calculation efficiency of the present embodiment, the preset conditions further include: the position belongs to and predetermines the position scope, and is concrete, predetermines the position scope and includes: and taking a router to which the WiFi data belongs as a center, and taking a preset distance as a radius to form a circular area, wherein the preset distance is the length of a straight line which can be covered by a signal of the router.

If the face data of the face does not have the face data group meeting the preset condition, the matching score between the WiFi data and the face is set to 0, and the following action of step a2 is not needed, and of course, the action of step a2 may be continuously executed.

And A2, calculating the matching score between each piece of WiFi data and the associated face data group to obtain the matching score of each piece of WiFi data.

In this step, the calculation process of the matching score between any piece of WiFi data and the associated face data group is the same, and for convenience of description, taking any piece of WiFi data and the associated face data group as an example, the calculation process of the matching score between the piece of WiFi data and the associated face data group is introduced, and specifically includes:

respectively calculating the matching scores between the WiFi data and the face data in the associated face data group, adding the calculated matching scores, if the value obtained by adding is not larger than a preset matching score threshold value, taking the value obtained by adding as the matching score between the WiFi data and the associated face data group, and if the value obtained by adding is larger than the preset matching score threshold value, taking the preset matching score threshold value as the matching score between the WiFi data and the associated face data group. For convenience of description, the matching score between the piece of WiFi data and the associated face data set is referred to as the matching score of the piece of WiFi data.

Wherein, the process of calculating the matching score between the WiFi data and any one of the face data in the associated face data group comprises:

firstly, according to the signal intensity in the WiFi data and the preset relationship between the signal intensity and the distance, the distance between the device to which the MAC address included in the WiFi data belongs and the router to which the WiFi data belongs is calculated to be a first distance. Secondly, determining that the distance between the face contained in the piece of face data and the router to which the piece of WiFi data belongs is a second distance. And finally, determining a matching score between the WiFi data and the face data according to a difference between a first distance and a second distance, wherein the difference between the first distance and the second distance is in negative correlation with the matching score between the WiFi data and the face data, namely the smaller the difference between the first distance and the second distance is, the higher the matching score between the WiFi data and the face data is, and conversely, the lower the matching score between the WiFi data and the face data is.

The matching score of each piece of WiFi data in the WiFi data group can be obtained through the step.

A3, determining the matching score between the MAC address contained in the WiFi data group and the face according to the matching score of each piece of WiFi data.

In this step, the calculation may be performed in two ways, namely, a first way and a second way.

Wherein, the first mode includes: and adding the matching scores of each piece of WiFi data in the WiFi data group, wherein the obtained value is used as the matching score between the MAC address contained in the WiFi data group and the face.

In practice, due to the WiFi data belonging to different routers in the WiFi data group and the WiFi data with overlapping time periods in the WiFi data belonging to different routers, that is, the noise data exists in the WiFi data group, the accuracy of the matching score between the WiFi data group and the face determined according to the first method cannot achieve a good effect. For example, the WiFi data belonging to the router a exists in the WiFi data group, and the time period is 9: 00 to 9: 30, and meanwhile, the WiFi data belonging to the router B exists in the WiFi data group, and the time period is 9: 20 to 9: 40, where 9: 20 to 9: 30 are overlapped time periods.

In order to further improve the accuracy of the matching score between the MAC address contained in the WiFi data set and the face, this step provides a second approach. Specifically, the process of calculating the matching score between the MAC address included in the WiFi data set and the face through the second method includes steps B1 to B3:

and B1, using the WiFi data with the time periods overlapping in the WiFi data belonging to different routers in the WiFi data group as a group of WiFi data to be processed.

Taking the overlapping time period of 20 points to 9 points 30 points of 9 as an example, in this step, the WiFi data that is in the WiFi data group and belongs to 20 points to 9 points 30 points of 9 in the WiFi data belonging to the router a, and the WiFi data that is in the WiFi data group and belongs to 20 points to 9 points 30 points of 9 in the WiFi data belonging to the router B are taken as a group of WiFi data to be processed.

And B2, weighting the matching scores of the WiFi data to be processed in each group of WiFi data to be processed.

Specifically, taking any group of WiFi data as an example, the moments of the WiFi data in the group of WiFi data to be processed are sorted according to a preset sequence, so as to obtain sorted WiFi data to be processed. And determining the weight of the matching score of each piece of the sorted WiFi data to be processed, wherein the weight of the matching score of any piece of the WiFi data is negatively related to a target distance, and the target distance is the distance between the router to which the piece of the WiFi data and the adjacent piece of the WiFi data belong respectively.

However, in the process of calculating the matching score of a group of WiFi data to be processed, for each piece of WiFi data to be processed, an adjacent piece of WiFi data is "previous piece of WiFi data", or an adjacent piece of WiFi data is "next piece of WiFi data", specifically, "previous piece of WiFi data" or "next piece of WiFi data" is not limited in this embodiment, as long as the same group of WiFi data to be processed is uniform.

Specifically, if the distance between the routers to which the respective routers belong is greater than a preset distance threshold, the weight may be set to a value greater than 1, and the greater the distance, the greater the weight. If the distance between the routers respectively belonging to the router is smaller than a preset distance threshold, the weight is set to be a numerical value smaller than 1, and the smaller the distance is, the smaller the weight is. For the first WiFi data (in the case of comparing the previous WiFi data) in any group of WiFi data to be processed, which is sorted from time to time, the weight may be set to 1. For the last WiFi data (in case of comparing the latter WiFi data) in the group of WiFi data to be processed sorted in time, the weight may be set to 1.

And B3, adding the matching scores of the WiFi data except the WiFi data to be processed in each group in the WiFi data group and the matching scores corresponding to the WiFi data to be processed in each group respectively, and taking the value obtained by adding as the matching score between the MAC address contained in the WiFi data group and the face.

And S303, taking the faces respectively corresponding to the preset number of matching scores which are sorted from large to small according to the matching scores of the target MAC address and the faces respectively as the faces matched with the target MAC address.

The target MAC address is a MAC address included in any WiFi data group, and is described by taking any WiFi data group as an example. Specifically, the faces corresponding to the preset number of matching scores are determined according to the sequence from high to low between the MAC addresses included in the WiFi data group and the respective faces, and are used as the faces matched with the MAC addresses included in the WiFi data group.

Based on this, the resulting matching data set may include: any WiFi data group comprises a MAC address and any human face matched with the WiFi data group.

In this step, the value of the preset number may be determined according to an actual service scenario, and the value of the preset number is not limited in this embodiment. In this embodiment, the value of the preset number may be 5.

For example, the MAC address included in the WiFi data group is a, and the face data matched with the WiFi data group includes a face: the method comprises the following steps of a face A, a face B, a face C and a face D, wherein the matching score between the face A and the face A is 80 points, the matching score between the face A and the face B is 90 points, the matching score between the face A and the face C is 90 points, the matching score between the face A and the face D is 100 points, and the value of the preset number is 3. The resulting set of matching data may be: a matching data group Z1{ MAC address A, face D }, a matching data group Z2{ MAC address A, face C }, and a matching data group Z3{ MAC address A, face B }.

Note that, there are the following cases: in order to improve the accuracy of the matching data set, the following S304 may be further performed, where the same MAC address exists in the faces respectively matched with the MAC addresses in different preset time periods, as follows:

s304, under the condition that the same MAC address exists in the faces respectively matched with the MAC addresses in different preset time periods, determining the face which repeatedly appears in the faces matched with the same MAC address, and taking the face as the face matched with the same MAC address in the different preset time periods.

In this step, the different preset time periods may be different days, for example, day 20 in 6 months in 2019 and day 21 in 6 months in 2019 are two different preset time periods.

In this embodiment, faces respectively matched with each MAC address included in the WiFi data may be determined from the WiFi data and the face data within each preset time. In this step, under the condition that the same MAC address exists in the faces respectively matched with the MAC addresses in different preset time periods, the faces repeatedly appearing in the faces respectively matched with the same MAC addresses in different preset time periods are taken as the faces matched with the same MAC addresses.

For example, a face matching the MAC address included in the WiFi data group is obtained every day of 20 in 2019 in 6 th month, 21 in 2019 in 6 th month, 22 in 2019 in 6 th month, and 23 in 2019 in 6 th month. In this step, a face that appears repeatedly is determined from faces that match the MAC address included in the WiFi data group. And taking the determined repeated face as a face matched with the MAC address contained in the WiFi data group, wherein the obtained matching result has higher accuracy.

For example, a minired at market X with friend a, friend B on day 20 of 6 months in 2019, may result in a face matching the MAC address of the minired device being minired, friend a, and friend B. Pony red in 2019, 6, 21 with friend C, friend D in mall X, it may be found that the face matching the MAC address of the pony red device has pony red, friend C, and friend D. If the device carried by the X shop in the 20 th 6 th month in 2019 and the 21 st 6 th month in 2019 is the same device, the 20 th 6 th month in 2019 and the 21 st 6 th month in 2019 can be obtained, and the face repeatedly appearing in the face matched with the MAC address of the small red device is small red, so that the face matched with the MAC address of the small red device can be determined to be small red, and the result is consistent with the actual situation, so that the matching result of taking the face repeatedly appearing in the face matched with the MAC address as the face matched with the MAC address has higher accuracy.

According to the technical scheme, the WiFi data of the same MAC address in the WiFi data in the same partition are used as a group to obtain a plurality of WiFi data groups, the matching scores between the MAC address contained in each WiFi data group and each face in the face data are calculated, the matching scores between each MAC address and each face are obtained, and for each MAC address, the faces corresponding to the preset number of matching scores in the matching scores between the MAC address and each face in the sequence from large to small are used as the faces matched with the MAC address.

The matching score is positively matched with the degree that the equipment to which the MAC address belongs and the face respectively indicate the same person, so that the determined face matched with the MAC address has certain accuracy, and the matching degree of the MAC address and the face in the obtained matching data group is high on the basis.

The data association device provided by the embodiment of the present application is described below, and the data association device described below and the data association method described above may be referred to in correspondence with each other.

Referring to fig. 4, a schematic structural diagram of a data association apparatus provided in an embodiment of the present application is shown, and as shown in fig. 4, the apparatus may include:

a data obtaining module 401, configured to obtain WiFi data and face data, where the WiFi data includes an MAC address of a device connected to a WiFi network, and the face data includes a face image;

a subdivision module 402, configured to divide the WiFi data into at least two parts, where WiFi data of the at least two parts do not intersect and have the same data amount, and WiFi data of the devices located on the same floor are included in the same part;

a partitioning module 403, configured to divide WiFi data of a target portion into multiple partitions according to multiple preset signal strength thresholds, where the target portion is any one portion;

a matching module 404, configured to obtain matching data sets by determining face data matched with the WiFi data in each partition, where any one of the matching data sets includes a matched MAC address and a face;

a counting module 405, configured to count a total number of times that a target matching data set in the target portion appears in the partition of the target portion, where the target matching data set is any one of the matching data sets;

a scoring module 406, configured to use an operation result of a first numerical value and a second numerical value as a score of the target matching data set, where the first numerical value is a total number of times that the target matching data set appears in all the portions, and the second numerical value is a total number of all the portions including the target matching data set;

and the association module 407 is configured to obtain a face matched with the MAC address according to the score of each matching data group.

Optionally, the partitioning module is configured to divide WiFi data of the target portion into a plurality of partitions according to a plurality of preset signal strength thresholds, and includes: the partitioning module is specifically configured to:

Optionally, the scoring module is configured to use a result of an operation between the first numerical value and the second numerical value as a score of the target matching data set, and includes: the scoring module is specifically configured to:

Optionally, the matching module is configured to obtain a matching data group by determining face data matched with the WiFi data in each of the partitions, and includes:

the data group acquisition unit is used for taking WiFi data with the same MAC address in the WiFi data as a group to obtain a WiFi data group;

the matching score calculation unit is used for calculating matching scores between the MAC address contained in each WiFi data group and each face in the face image respectively, and the size of the matching score is positively correlated with the degree that the equipment to which the MAC address belongs and the face to which the face image belongs respectively indicate the same person;

and the face matching unit is used for taking faces respectively corresponding to a preset number of matching scores which are sorted from large to small according to the matching scores in the matching scores between the target MAC address and each face as the faces matched with the target MAC address, wherein the target MAC address is any one MAC address.

Optionally, the matching score calculating unit is configured to determine a matching score between a MAC address included in any WiFi data group and any face in the face image, and includes: the matching score calculating unit is specifically configured to:

respectively calculating the matching score between each piece of WiFi data and the associated face data group to obtain the matching score of each piece of WiFi data;

Optionally, the matching score calculating unit is configured to calculate a matching score between any piece of WiFi data and the associated face data set, and includes: the matching score calculating unit is specifically configured to:

and adding the WiFi data and the matching score of each piece of face data in the associated face data group to obtain a value, and taking the value as the matching score between the WiFi data and the associated face data group.

Optionally, the matching score calculating unit is further specifically configured to: after the matching score between each piece of WiFi data and the associated face data group is calculated respectively to obtain the matching score of each piece of WiFi data, and before the matching score between the MAC address contained in the WiFi data group and the face is determined according to the matching score of each piece of WiFi data:

optionally, the matching score calculating unit is specifically configured to:

An embodiment of the present application further provides a data association device, please refer to fig. 5, which shows a schematic structural diagram of the data association device, where the data association device may include: at least one processor 501, at least one communication interface 502, at least one memory 503, and at least one communication bus 504;

in this embodiment, the number of the processor 501, the communication interface 502, the memory 503 and the communication bus 504 is at least one, and the processor 501, the communication interface 502 and the memory 503 complete communication with each other through the communication bus 504;

the processor 501 may be a central processing unit CPU, or an application Specific Integrated circuit asic, or one or more Integrated circuits configured to implement embodiments of the present invention, etc.;

the memory 503 may include a high-speed RAM memory, and may further include a non-volatile memory (non-volatile memory) or the like, such as at least one disk memory;

wherein the memory stores a program and the processor can call the program stored in the memory, the program for:

Alternatively, the detailed function and the extended function of the program may be as described above.

Embodiments of the present application further provide a readable storage medium, where a program suitable for being executed by a processor may be stored, where the program is configured to:

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 data association method, comprising:

obtaining matched data groups by determining face data matched with WiFi data in each partition, wherein any one matched data group comprises an MAC address and a face matched with WiFi;

and obtaining the face matched with the WiFi of the MAC address according to the scores of the matched data groups.

2. The method of claim 1, wherein the dividing WiFi data of the target portion into a plurality of partitions according to a plurality of preset signal strength thresholds comprises:

3. The method of claim 2, wherein the signal strength threshold is determined according to a distribution of signal strengths of the WiFi network-connected devices and/or distances between the WiFi access points and the devices.

4. The method of claim 1, wherein the operation of the first numerical value and the second numerical value as the score of the target matching data set comprises:

5. The method according to any one of claims 1 to 4, wherein the determining face data of the WiFi data match in each of the partitions to obtain a matching data set comprises:

calculating matching scores between the MAC address contained in each WiFi data group and each face in the face image respectively, wherein the size of the matching score is positively correlated with the degree that equipment to which the MAC address belongs and the degree that the face to which the face image belongs respectively indicate the same person;

6. The method of claim 5, wherein determining a matching score between a MAC address contained in any one of the WiFi data sets and any one of the faces in the face image comprises:

determining face data meeting preset conditions with each WiFi data in the WiFi data group from the face data of the face respectively, and taking the face data as a face data group associated with the WiFi data; for any piece of WiFi data, the preset conditions include: the time belongs to a preset time range; the preset time range is a time range formed by preset time lengths before and after the generation time of the WiFi data is taken as a time midpoint;

7. The method of claim 6, wherein calculating a match score between any one of the WiFi data and the associated face data set comprises:

calculating the matching score of the WiFi data and each piece of face data in the associated face data group; the matching score of any piece of face data in the WiFi data and the associated face data group is in negative correlation with the difference; the difference is a difference between the first distance and the second distance; the first distance is the distance between the position in the WiFi data and the router to which the WiFi data belongs; the second distance is the distance between the position in the piece of face data and the router to which the piece of WiFi data belongs;

8. The method of claim 6, wherein after the separately calculating the matching score between each WiFi data and the associated face data group to obtain the matching score of each WiFi data, and before the determining the matching score between the MAC address included in the WiFi data group and the face according to the matching score of each WiFi data, further comprises:

9. A data association apparatus, comprising:

10. A data association device, comprising: a memory and a processor;

the memory is used for storing programs;

the processor is used for executing the program and realizing the steps of the data association method according to any one of claims 1 to 8.

11. A readable storage medium, having stored thereon a computer program, wherein the computer program, when executed by a processor, performs the steps of the data correlation method according to any one of claims 1 to 8.