CN105279496B - Method and device for face recognition - Google Patents

Abstract

The present invention provides a method and device for face recognition. The method includes: obtaining the first user equipment information collected by front-end equipment and the first facial feature information of the face image; maintaining the second user equipment information in advance with the second The association relationship between the face feature information; searching the second user device information for information that matches the multiple first user device information within the period of collecting the first face feature information, and finding the information associated with the multiple first user device information The second facial feature information: selecting the second facial feature information corresponding to the first facial feature information from the second facial feature information associated with the plurality of first user equipment information, and determining the face recognition result. Through the technical solution of the present invention, the management server does not determine the face recognition result only based on the face feature information of the face image collected by the front-end equipment, and does not completely depend on the face feature information of the face image, so that the face recognition can be improved. Improve the recognition accuracy and improve the reliability of face recognition.

Description

Method and device for face recognition

technical field

The present invention relates to the field of video technology, in particular to a face recognition method and device.

Background technique

In recent years, with the rapid development of computers, networks, image processing, and transmission technologies, the trend of popularization of video surveillance has become more and more obvious. Video surveillance is gradually becoming high-definition and intelligent. Video surveillance systems can be used in many fields. Such as intelligent transportation, smart park, safe city, etc.

With the rapid development of computer image recognition technology, in video surveillance systems, face recognition technology has more application requirements, such as face capture at entrances and exits, face attendance, personnel control, etc. Face provides a wealth of human Facial feature information can play an important role in the process of person recognition.

In the current face recognition technology, face recognition can be performed based on video images containing faces. Specifically, extract the face feature information from the video image containing the face, and compare the extracted face feature information with the face feature information stored in the face database to obtain the similarity ranking of the faces, And select the face with the highest similarity from the face database as the recognition result for output.

The above method completely depends on the face feature information of the video image, and the video image has high requirements on the environment, is affected by light and angle, has poor anti-interference ability, and has a high rate of missed recognition and misjudgment in face recognition, and it is even impossible to distinguish whether it is a human face or not. Real faces, such as face images on clothes, face reflections on glass doors, etc.

Contents of the invention

The present invention provides a method for face recognition, which is applied to a management server. The method includes the following steps: obtaining the first user equipment information collected by the front-end device and the first face feature information of the face image; maintaining the first face feature information in advance 2. Association relationship between user equipment information and second facial feature information; search information from second user equipment information that matches multiple first user equipment information within the period of collecting first facial feature information, and find multiple The second facial feature information associated with the first user equipment information; selecting the second human face corresponding to the first facial feature information from the second facial feature information associated with the plurality of first user equipment information Feature information to determine the face recognition result.

Before searching the second user equipment information for information that matches a plurality of first user equipment information within the period of collecting the first facial feature information, and finding the second facial feature information associated with the multiple first user equipment information, The method also includes: judging whether there is second facial feature information whose similarity with the first facial feature information meets the preset requirements in all the second facial feature information; The second face feature information whose similarity between the face feature information meets the preset requirements is used as the face recognition result; if not, perform a search from the second user device information and collect multiple face feature information within the period of time. The process of finding the second facial feature information associated with multiple first user equipment information based on the information matched by the first user equipment information.

The process of selecting the second facial feature information corresponding to the first facial feature information from the second facial feature information associated with the plurality of first user equipment information, and determining the face recognition result, specifically The method includes: comparing the first facial feature information with each second facial feature information associated with the plurality of first user equipment information to obtain each of the plurality of first user equipment information associated The similarity between the second facial feature information and the first facial feature information; the second facial feature information with the highest similarity is selected as the face recognition result.

The second user equipment information includes one or any combination of the following: wearable information, the serial number of the mobile terminal, the MAC address of the mobile terminal, the number of the mobile terminal, and the serial number of the SIM card; the first user equipment information includes The content is the same as that included in the second user equipment information.

The present invention provides a device for face recognition, which is applied to a management server. The device specifically includes: an obtaining module, which is used to obtain the first user device information collected by the front-end device, and obtain the person information collected by the front-end device. The first face feature information corresponding to the face image; the maintenance module is used to pre-maintain the association relationship between the second user equipment information and the second face feature information; the query module is used to search for information from the second user equipment information The information matched with the multiple first user equipment information within the period of collecting the first facial feature information is used to find the second facial feature information associated with the multiple first user equipment information; The second facial feature information corresponding to the first facial feature information is selected from the second facial feature information associated with the first user equipment information, and the face recognition result is determined.

The query module is further configured to search the second user equipment information for information that matches a plurality of first user equipment information within the period of collecting the first facial feature information, and find the first user equipment information associated with the plurality of first user equipment information. Before the feature information of two people's faces, it is judged whether there is a second face feature information whose similarity with the first face feature information meets the preset requirements in all the second face feature information; if not, then execute The process of searching the second user equipment information for information that matches the multiple first user equipment information within the period of collecting the first facial feature information, and finding the second facial feature information associated with the multiple first user equipment information; The determination module is further configured to use the second facial feature information whose similarity with the first facial feature information meets the preset requirements as the face recognition result when the query result is yes.

The determination module is specifically configured to filter out the second facial feature information corresponding to the first facial feature information from the second facial feature information associated with the plurality of first user equipment information, and determine the person In the process of face recognition results, the first facial feature information is compared with each second facial feature information associated with the multiple first user equipment information to obtain the multiple first user The similarity between each second facial feature information associated with the device information and the first facial feature information; select the second facial feature information with the highest similarity as the face recognition result.

The second user equipment information includes one or any combination of the following: wearable information, the serial number of the mobile terminal, the MAC address of the mobile terminal, the number of the mobile terminal, and the serial number of the SIM card; the first user equipment information includes The content is the same as that included in the second user equipment information.

Based on the above technical solution, in the embodiment of the present invention, the face recognition result can be determined based on the user equipment information collected by the front-end device and the face feature information of the face image, instead of only based on the face image collected by the front-end device. Face feature information, to determine the result of face recognition, does not completely depend on the face feature information of the face image, so as to determine the result of face recognition through multi-dimensional information, improve the accuracy of face recognition, and reduce the rate of missed recognition and errors improve the reliability of face recognition.

Description of drawings

Fig. 1 is the flowchart of the method for face recognition in an embodiment of the present invention;

Fig. 2 is the flow chart of the method for face recognition in another embodiment of the present invention;

FIG. 3 is a hardware structural diagram of a management server in an embodiment of the present invention;

Fig. 4 is a structural diagram of a face recognition device in an embodiment of the present invention.

Detailed ways

Aiming at the problems existing in the prior art, a method for face recognition is proposed in the embodiment of the present invention, which can be applied to a management server (such as a VM (Video Management, video management) server) of a video surveillance system, as shown in FIG. 1, the face recognition method may include the following steps:

Step 101, obtaining first user equipment information collected by a front-end device and first facial feature information of a facial image. Wherein, the front-end device may collect the first user device information and face image, and send the first user device information and face image to the management server. The management server obtains the first user equipment information collected by the front-end device based on the first user equipment information from the front-end device. Based on the face image from the front-end device, the management server obtains the first face feature information corresponding to the face image, and the first face feature information is the first face feature information of the face image collected by the front-end device.

Step 102, maintaining the association relationship between the second user equipment information and the second facial feature information in advance.

Wherein, the association relationship between the second user equipment information and the second facial feature information may specifically include: the correspondence between the second user equipment information and personnel information, and the correspondence between the second facial feature information and personnel information. Correspondence; in the process of pre-maintaining the correspondence between the second user equipment information and the personnel information, and the correspondence between the second facial feature information and the personnel information, the second user equipment information and personnel information can be obtained, and Store the corresponding relationship between the second user equipment information and the currently obtained personnel information in the pre-maintained device information database; obtain the second facial feature information and personnel information of the face image, and store them in the pre-maintained feature information database The corresponding relationship between the second facial feature information and the currently obtained personnel information is stored. Or, obtain the second user device information, the second facial feature information and personnel information of the face image, and store the second user device information, the second facial feature information and the currently obtained personnel information in the pre-maintained feature information database Correspondence between information.

Step 103, look for information from the second user equipment information that matches a plurality of first user equipment information within the period of collecting the first facial feature information (the time interval in which the front-end equipment collects the facial image) (that is, the second user equipment information) to find the second facial feature information associated with multiple pieces of first user equipment information (the second facial feature information associated with multiple second user equipment information corresponding to multiple first user equipment information).

Step 104: Select the second facial feature information corresponding to the first facial feature information from the second facial feature information associated with the plurality of first user equipment information, and determine the face recognition result. Wherein, the face recognition result is the face corresponding to the face image collected by the front-end device, and the corresponding personnel information can be output.

In the embodiment of the present invention, after searching the second user equipment information for information that matches the multiple first user equipment information within the period of collecting the first facial feature information, the second person associated with the multiple first user equipment information is found. Before the face feature information, it is also possible to judge all the second face feature information (that is, all the second face feature information stored in the feature information database, which will include the second face feature information corresponding to the first user equipment information) Among them, whether there is a second face feature information whose similarity with the first face feature information meets the preset requirements; if yes, the similarity with the first face feature information can be directly satisfied The required second face feature information is used as the face recognition result; if not, then perform a search from the second user device information for information that matches a plurality of first user device information within the period of collecting the first face feature information, and find A process of associating multiple pieces of first user equipment information with second facial feature information.

In the embodiment of the present invention, the information (that is, the second user equipment information) that matches the multiple first user equipment information within the period of collecting the first facial feature information is searched from the second user equipment information, and multiple first user equipment information is found. The process of associating the second facial feature information with the device information may specifically include but not limited to the following methods: use multiple first user device information within the period of collecting the first facial feature information to query the pre-maintained second user device information and Correspondence between personnel information, find out the second user equipment information matching the plurality of first user equipment information from the second user equipment information, and obtain the personnel information corresponding to these second user equipment information; use the obtained The personnel information queries the correspondence between the pre-maintained second facial feature information and the personal information, and obtains a plurality of corresponding second facial feature information, and the obtained second facial feature information is a plurality of first user equipment information Associated second facial feature information.

In the embodiment of the present invention, during the time interval when the front-end device collects the face image, the front-end device collects a plurality of first user device information, and the plurality of first user device information corresponds to a plurality of second face feature information. Based on this, the process of selecting the second facial feature information corresponding to the first facial feature information from the second facial feature information associated with multiple first user equipment information, and determining the face recognition result may specifically include but not The method is limited to the following method: compare the first facial feature information with each second facial feature information associated with the plurality of first user equipment information corresponding to the time interval in which the facial image is collected, and obtain the plurality of first facial features. The similarity between each second facial feature information associated with the user equipment information and the first facial feature information; select the second facial feature information with the highest similarity as the face recognition result.

In the above process of the embodiment of the present invention, the second user equipment information may specifically include but not limited to one or any combination of the following: wearable information (such as Bluetooth bracelet information, Bluetooth headset information, etc.), serial number of the mobile terminal, mobile The MAC (Media Access Control, Media Access Control) address of the terminal, the number of the mobile terminal, the SIM (Subscriber Identity Module, customer identification module) card serial number, etc.; the content included in the first user equipment information and the content included in the second user equipment information The content is the same.

For example, when the second user equipment information is wearable information (such as Bluetooth bracelet information, Bluetooth headset information, etc.), the first user equipment information collected by the front-end device is wearable information (such as Bluetooth bracelet information, Bluetooth headset information, etc.). information, etc.). When the second user equipment information is the MAC address of the mobile terminal, the first user equipment information collected by the front-end device is the MAC address of the mobile terminal.

Of course, in practical applications, both the first user equipment information and the second user equipment information are not limited to the above user equipment information, and details about other user equipment information will not be repeated here.

Based on the above technical solution, in the embodiment of the present invention, the face recognition result can be determined based on the user equipment information collected by the front-end device and the face feature information of the face image, instead of only based on the face image collected by the front-end device. Face feature information, to determine the result of face recognition, does not completely depend on the face feature information of the face image, so as to determine the result of face recognition through multi-dimensional information, improve the accuracy of face recognition, and reduce the rate of missed recognition and errors improve the reliability of face recognition.

The above process will be further described below in conjunction with specific embodiments.

In the embodiment of the present invention, a face recognition method is proposed, which can be applied in a video surveillance system. The video surveillance system can at least include a front-end device (such as a camera) and a management server (such as a VM server), as shown in FIG. 2 As shown, the method for face recognition may specifically include the following steps:

In step 201, the management server obtains the second user equipment information and personnel information, and stores the correspondence between the second user equipment information and the personnel information in a pre-maintained equipment information database.

Wherein, the second user equipment information includes, but is not limited to, one or any combination of the following: wearable information of the person (such as Bluetooth bracelet information, Bluetooth headset information, etc.), mobile terminal information of the person (such as the serial number of the mobile terminal, mobile MAC address of the terminal, number of the mobile terminal, serial number of the SIM card, etc.).

Wherein, the management server may store the corresponding relationship between the personal wearable information, the personal mobile terminal information and the personal information in the same device information database, for example, store the personal wearable information, personal Correspondence between mobile terminal information and personnel information. The management server can also store the correspondence between personal wearable information and personnel information, and the correspondence between personnel's mobile terminal information and personnel information in different device information databases. The corresponding relationship between the personal wearable information and the personal information, and the corresponding relationship between the personal mobile terminal information and the personal information is stored in the device information library 2 .

Step 202, the management server obtains the second facial feature information and person information of the face image, and stores the correspondence between the second facial feature information and the person information in a pre-maintained feature information database. Wherein, the face image may specifically refer to a video image including a face.

Wherein, the second face feature information may include the face image, and may also include other information of the face, such as the distance between two eyes, the distance between eyebrows, facial features, etc., which will not be repeated here.

In a specific application scenario, the second user equipment information, the second facial feature information, and the personnel information may be obtained in a corresponding manner. For example, for a park, people in the park can be required to provide their own personnel information (such as name, ID card, gender, age, etc.) Personnel collection, clarity, accuracy are very high).

For steps 201 and 202, the management server stores the correspondence between the second facial feature information and the personnel information, the correspondence between the second user equipment information and the personnel information in the characteristic information database and the equipment information database respectively, In practical applications, the management server may also store the correspondence between the second user equipment information, the second facial feature information, and the currently obtained personnel information in the same information database (such as a feature information database). No more details.

Step 203, the management server obtains the first user equipment information and the first facial feature information of the facial image collected by the front-end equipment. Wherein, the front-end device may collect the first user device information and the face image, and send the first user device information and the face image to the management server. Further, the management server may obtain the first user equipment information collected by the front-end device based on the first user equipment information from the front-end device. The management server may also obtain the first face feature information corresponding to the face image based on the face image from the front-end device, and the first face feature information is the first face feature information corresponding to the face image collected by the front-end device. Facial feature information.

In the embodiment of the present invention, for the convenience of distinction, the face feature information stored in the feature information database is called the second face feature information, and the face feature information corresponding to the face image collected by the front-end device is called the first face feature information. Facial feature information. The user equipment information stored in the equipment information database is called second user equipment information, and the user equipment information collected by the front-end device is called first user equipment information.

Among them, through the image sensor of the front-end equipment (such as a camera) installed in a specific area of the key entrance and exit, the face image can be captured, and the front-end equipment sends the face image to the management server, and the management server obtains the information collected by the front-end equipment. The first face feature information of the face image.

Wherein, through the sensing detection module (which can be installed on the front-end equipment, such as WIFI (Wireless Fidelity, wireless fidelity) wireless module, bluetooth module, etc.) The first user equipment information is collected, and the front-end device sends the first user equipment information to the management server, and the management server obtains the first user equipment information collected by the front-end device, and the content included in the first user equipment information is consistent with the second The user equipment information includes the same content. For example, when the second user equipment information is wearable information (such as Bluetooth bracelet information, Bluetooth headset information, etc.), the first user equipment information collected by the front-end device is wearable information (such as Bluetooth bracelet information, Bluetooth headset information, etc.). information, etc.). When the second user equipment information is the MAC address of the mobile terminal, the first user equipment information collected by the front-end device is the MAC address of the mobile terminal. Of course, in practical applications, both the first user equipment information and the second user equipment information are not limited to the above user equipment information, and details about other user equipment information will not be repeated here.

Wherein, since there may be multiple users located in the same area, at each collection moment, the first user equipment information collected by the front-end device may be a plurality of first user equipment information.

Wherein, when the front-end device sends the collected face image to the management server, it also sends the collection time of the face image to the management server, and the management server records the corresponding relationship between the first facial feature information of the face image and the collection time . When the front-end device sends the collected first user equipment information to the management server, it may also send the collection time of the first user equipment information to the management server, and the management server records the correspondence between the first user equipment information and the collection time.

Among them, taking the first user equipment information collected by the WIFI wireless module of the front-end equipment as an example, the WIFI wireless module periodically sends Beacon (detection) messages, and when the mobile terminal is within the detection range of the WIFI wireless module, the mobile terminal will send The WIFI wireless module returns a response message, and the WIFI wireless module can obtain the first user equipment information from the response message returned by the mobile terminal, and obtain the collection time of the first user equipment information. Alternatively, the mobile terminal actively and periodically sends a detection request message. When the mobile terminal is located within the detection range of the WIFI wireless module, the detection request message sent by the mobile terminal will be received by the WIFI wireless module, and the WIFI wireless module can receive data from the mobile terminal. Obtain the first user equipment information from the detection request message, and obtain the collection time of the first user equipment information.

Wherein, when the front-end device sends the collected first user equipment information to the management server, within a configurable time period T, the front-end device only sends the first user equipment information to the management server once.

Step 204, the management server searches the information of the second user equipment for information that matches the multiple first user equipment information within the period of collecting the first facial feature information (that is, the time interval during which the front-end equipment collects the facial image), and finds multiple The second facial feature information associated with a piece of first user equipment information (the second facial feature information associated with a plurality of second user equipment information corresponding to a plurality of first user equipment information).

In the embodiment of the present invention, the management server searches the second user equipment information for information that matches the multiple first user equipment information (that is, the second user equipment information) within the period of collecting the first facial feature information, and finds multiple first user equipment information. The process of associating the second facial feature information with user equipment information may specifically include but not limited to the following manner: the management server uses multiple first user equipment information within the period of collecting the first facial feature information to query the pre-maintained second facial feature information. The corresponding relationship between user equipment information and personnel information, the management server finds the second user equipment information that matches the plurality of first user equipment information from the second user equipment information, and obtains the information corresponding to the second user equipment information Personnel information; the management server uses the obtained personnel information to query the correspondence between the second human face feature information maintained in advance and the personnel information, and obtains a plurality of corresponding second human face feature information, and the second human face obtained by the management server The feature information is second face feature information associated with multiple pieces of first user equipment information.

In step 205, the management server selects the second facial feature information corresponding to the first facial feature information from the second facial feature information associated with the plurality of first user equipment information, and determines the face recognition result. Wherein, the face recognition result is the face corresponding to the face image collected by the front-end device, and the management server may also output the personnel information corresponding to the face recognition result after obtaining the face recognition result.

Among them, in the time interval when the front-end device collects the face image, the front-end device collects a plurality of first user device information, and the management server uses the plurality of first user device information to query the relationship between the pre-maintained second user device information and personnel information. Correspondence between the corresponding relationship, obtain a plurality of personnel information, and use the plurality of personnel information to query the corresponding relationship between the pre-maintained second facial feature information and personnel information, and obtain a plurality of second facial feature information, that is, multiple first facial feature information The user equipment information corresponds to a plurality of second facial feature information.

Based on this, in the embodiment of the present invention, the management server selects the second face feature information corresponding to the first face feature information from the second face feature information (multiple second face feature information) associated with the first user equipment information. The facial feature information, the process of determining the face recognition result may specifically include but not limited to the following manner: the management server stores the first facial feature information and the plurality of first user device information corresponding to the time interval when the facial image is collected Each associated second face feature information is compared to obtain the similarity between each second face feature information associated with the plurality of first user equipment information and the first face feature information; the management server selects the highest similarity The second face feature information is used as the face recognition result.

Wherein, after the management server obtains the first facial feature information of the facial image, it determines the collection time of the facial image, and in consideration of the influence of factors such as errors, it can also determine the time interval for collecting the facial image (that is, including the time interval of the facial image). The time interval of the time when the face image is collected), and, in the time interval when the face image is collected, the management server may determine a plurality of pieces of first user equipment information.

In the embodiment of the present invention, the management server searches the second user equipment information for information that matches the multiple first user equipment information within the period of collecting the first facial feature information, and finds the first user equipment information associated with the multiple first user equipment information. Before the feature information of two faces, the management server can also judge all the second face feature information (that is, all the second face feature information stored in the feature information database, which will include the second person corresponding to the first user equipment information) In the face feature information), whether there is second face feature information whose similarity with the first face feature information meets the preset requirements. If so, the management server may directly use the second facial feature information whose similarity with the first facial feature information meets the preset requirements as the face recognition result, and the above step 204 and step 205 are not performed again. If not, the management server performs a search from the second user equipment information for information that matches multiple first user equipment information within the period of collecting the first facial feature information, and finds the second person associated with multiple first user equipment information The process of facial feature information.

Based on the above method, in the embodiment of the present invention, the management server first judges whether there is second facial feature information whose similarity with the first facial feature information meets the preset requirements among all the second facial feature information. . When the first face feature information of the video image is relatively clear, then all the second face feature information will have the second face feature information whose similarity with the first face feature information meets the preset requirements, so The second facial feature information whose similarity with the first facial feature information satisfies the preset requirements can be directly used as the face recognition result, so as to accurately obtain the face recognition result. For example, when clear first facial feature information 1 is obtained, even if the second facial feature information corresponding to the first user equipment information is second facial feature information 1, second facial feature information 2, second facial feature information feature information 3, then the management server also first judges whether there is a gap between the first face feature information 1 and the first face feature information in all the second face feature information (such as the second face feature information 1-the second face feature information 1000). The second facial feature information whose similarity meets the preset requirements, assuming that the similarity between the second facial feature information 100 and the first facial feature information meets the preset requirements, then the second facial feature information 100 is used as The face recognition result, the face recognition result will be an accurate face recognition result. At this time, the management server does not need to obtain the second face feature information 1, the second face feature information 2, the second face feature information In step 3, the face recognition result corresponding to the first face feature information 1 is selected, and an accurate face recognition result can be obtained.

Wherein, after the management server obtains the first face feature information of the face image, it can also analyze the face image to detect whether the face image satisfies the face recognition requirements, such as whether the light of the face image conforms to Requirements, whether the angle meets the requirements, whether the distance between the two eyes meets the requirements, etc., the management server can use the Adaboost classifier to detect whether the face image meets the face recognition requirements, and the specific detection process will not be described in detail. If the face image does not meet the face recognition requirements, the management server compares the first face feature information with each second face feature information to obtain each second face feature information and the first face feature information , and select the second face feature information with the highest similarity as the face recognition result. Alternatively, the management server directly uses each second face feature information as a face recognition result, and at this time there will be multiple face recognition results. When the management server compares the first face feature information with each second face feature information, the number of feature points for comparison can be reduced. For example, when comparing normal face feature information, 100 feature points are required , when the face image does not meet the face recognition requirements, only 20 feature points are used when comparing face feature information.

In a specific application, if the face image meets the face recognition requirements, the management server combines the first face feature information with all the second face feature information in the feature information database (such as the second face feature information 1- The second facial feature information (1000) is compared to obtain the similarity between each second facial feature information in the feature information base and the first facial feature information, if any similarity exceeds the set threshold (illustrating the similarity The corresponding second face feature information is the second face feature information whose similarity with the first face feature information meets the preset requirements), then the second face feature information with the highest similarity is selected as the face Recognition results, no longer perform the above step 204 and step 205; if no similarity exceeds the set threshold, each second facial feature information corresponding to the first user equipment information (such as the second Face feature information 1, the second face feature information 2, and the second face feature information 3) are compared to obtain each second face feature information corresponding to the first user equipment information and the first face feature information similarity, and select the second face feature information with the highest similarity as the face recognition result.

Wherein, when comparing two face feature information to obtain the similarity of two face feature information, a similarity algorithm can be used for processing, for example, the similarity algorithm can perceive hash algorithm, SIM (Structural SIMilarity, structural similarity ) algorithm, etc., which will not be described in detail here.

Based on the above technical solution, in the embodiment of the present invention, the face recognition result can be determined based on the user equipment information collected by the front-end device and the face feature information of the face image, instead of only based on the face image collected by the front-end device. Face feature information, to determine the result of face recognition, does not completely depend on the face feature information of the face image, so as to determine the result of face recognition through multi-dimensional information, improve the accuracy of face recognition, and reduce the rate of missed recognition and errors improve the reliability of face recognition.

Based on the same inventive concept as the above method, an embodiment of the present invention also provides a face recognition device, and the face recognition device is applied on a management server. Wherein, the device for face recognition can be implemented by software, or by hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, it is formed by reading the corresponding computer program instructions in the non-volatile memory into the memory for operation through the processor of the management server where it is located. From the hardware level, as shown in Figure 3, it is a hardware structural diagram of the management server where the face recognition device proposed by the present invention is located, except for the processor, network interface, memory and non-volatile memory shown in Figure 3 In addition to the non-volatile memory, the management server may also include other hardware, such as forwarding chips responsible for processing packets, etc.; in terms of hardware structure, the management server may also be a distributed device, which may include multiple interface cards, so as to implement Extensions for message handling.

As shown in FIG. 4 , it is a structural diagram of the device for face recognition proposed by the present invention. The device specifically includes: an obtaining module 11 for obtaining the first user device information collected by the front-end device, and obtaining the first user device information collected by the front-end device. The first face feature information corresponding to the collected face image; the maintenance module 12 is used to pre-maintain the association relationship between the second user equipment information and the second face feature information; the query module 13 is used to obtain from the first In the second user equipment information, search for information that matches a plurality of first user equipment information in the period of collecting the first facial feature information, and find the second facial feature information associated with a plurality of first user equipment information; the determination module 14 uses The second facial feature information corresponding to the first facial feature information is selected from the second facial feature information associated with the plurality of first user equipment information to determine a face recognition result.

The query module 13 is further configured to search the second user equipment information for information that matches a plurality of first user equipment information within the period of collecting the first facial feature information, and find the information associated with the plurality of first user equipment information. Before the second facial feature information, it is judged whether there is a second facial feature information whose similarity with the first facial feature information meets the preset requirements in all the second facial feature information; if not, then Execute the process of searching the second user equipment information for information that matches the multiple first user equipment information within the period of collecting the first facial feature information, and finding the second facial feature information associated with the multiple first user equipment information; The determining module 14 is further configured to use the second facial feature information whose similarity with the first facial feature information satisfies a preset requirement as the face recognition result when the query result is yes.

The determining module 14 is specifically configured to filter out the second facial feature information corresponding to the first facial feature information from the second facial feature information associated with the plurality of first user equipment information, and determine In the process of face recognition results, the first facial feature information is compared with each second facial feature information associated with the plurality of first user equipment information to obtain the plurality of first user equipment information. The similarity between each second facial feature information associated with the user equipment information and the first facial feature information;

Select the second face feature information with the highest similarity as the face recognition result.

In the embodiment of the present invention, the second user equipment information includes one or any combination of the following: wearable information, serial number of the mobile terminal, media access control MAC address of the mobile terminal, number of the mobile terminal, customer identification module SIM card Serial number; the content included in the first user equipment information is the same as the content included in the second user equipment information.

Wherein, each module of the device of the present invention can be integrated into one body, or can be deployed separately. The above modules can be combined into one module, or can be further split into multiple sub-modules.

Through the description of the above embodiments, those skilled in the art can clearly understand that the present invention can be realized by means of software plus a necessary general-purpose hardware platform, and of course also by hardware, but in many cases the former is a better implementation Way. Based on this understanding, the essence of the technical solution of the present invention or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions to make a A computer device (which may be a personal computer, a server, or a network device, etc.) executes the methods described in various embodiments of the present invention. Those skilled in the art can understand that the drawing is only a schematic diagram of a preferred embodiment, and the modules or processes in the drawing are not necessarily necessary for implementing the present invention.

Those skilled in the art can understand that the modules in the device in the embodiment can be distributed in the device in the embodiment according to the description in the embodiment, or can be located in one or more devices different from the embodiment according to corresponding changes. The modules in the above embodiments can be combined into one module, and can also be further divided into multiple sub-modules. The serial numbers of the above embodiments of the present invention are for description only, and do not represent the advantages and disadvantages of the embodiments.

The above disclosures are only a few specific embodiments of the present invention, however, the present invention is not limited thereto, and any changes conceivable by those skilled in the art shall fall within the protection scope of the present invention.

Claims (8)

1. a kind of method of recognition of face is applied in management server, which is characterized in that the described method comprises the following steps:

Obtain the first face characteristic information of headend equipment collected first user equipment information and facial image；

Incidence relation between maintenance second user facility information and the second face characteristic information in advance；

Believe from being searched in second user facility information with multiple first user equipmenies in the first face characteristic information period of acquisition Matched information is ceased, associated second face characteristic information of multiple first user equipment informations is found；

It is special that first face is filtered out from associated second face characteristic information of multiple first user equipment informations Reference ceases corresponding second face characteristic information, determines face recognition result.

2. the method according to claim 1, wherein described search and acquisition the from second user facility information Multiple matched information of first user equipment information in one face characteristic information period, find multiple first user equipment informations Before associated second face characteristic information, the method also includes:

Judge to meet in advance in the second all face characteristic informations with the presence or absence of the similarity between the first face characteristic information If it is required that the second face characteristic information；If so, the similarity between the first face characteristic information is met preset requirement The second face characteristic information as face recognition result；It searches and acquires from second user facility information if not, executing Multiple matched information of first user equipment information in first face characteristic information period find multiple first user equipment letters Cease the process of associated second face characteristic information.

3. the method according to claim 1, wherein described be associated with from multiple first user equipment informations The second face characteristic information in filter out corresponding second face characteristic information of first face characteristic information, determine face The process of recognition result, specifically includes:

By first face characteristic information and associated each second face characteristic of multiple first user equipment informations Information is compared, and obtains associated each second face characteristic information of multiple first user equipment informations and described the The similarity of one face characteristic information；

Highest second face characteristic information of similarity is selected, as face recognition result.

4. method according to claim 1-3, which is characterized in that the second user facility information includes following One of or any combination: body-worn information, the sequence number of mobile terminal, mobile terminal MAC address, move Number, the client identification module SIM card sequence number of dynamic terminal；The content and described second that first user equipment information includes The content that user equipment information includes is identical.

5. a kind of device of recognition of face is applied in management server, which is characterized in that described device specifically includes:

Module is obtained, for obtaining collected first user equipment information of headend equipment, and obtains the headend equipment acquisition First face characteristic information corresponding to the facial image arrived；

Maintenance module, for safeguarding the incidence relation between second user facility information and the second face characteristic information in advance；

Enquiry module, for from being searched in second user facility information and multiple the in the first face characteristic information period of acquisition The matched information of one user equipment information finds associated second face characteristic information of multiple first user equipment informations；

Determining module, for filtering out institute from associated second face characteristic information of multiple first user equipment informations Corresponding second face characteristic information of the first face characteristic information is stated, determines face recognition result.

6. device according to claim 5, which is characterized in that

The enquiry module is also used to out of lookup in second user facility information with the first face characteristic information period of acquisition Multiple matched information of first user equipment information, find multiple first user equipment informations associated second face characteristic letter Before breath, judge in the second all face characteristic informations with the presence or absence of the similarity between first face characteristic information Meet the second face characteristic information of preset requirement；It is searched from second user facility information and acquisition the if it is not, then executing Multiple matched information of first user equipment information in one face characteristic information period, find multiple first user equipment informations The process of associated second face characteristic information；

The determining module is also used to when query result, which is, is, then will be similar between first face characteristic information Degree meets the second face characteristic information of preset requirement as face recognition result.

7. device according to claim 5, which is characterized in that

The determining module, specifically for from associated second face characteristic information of multiple first user equipment informations In filter out corresponding second face characteristic information of first face characteristic information, during determining face recognition result, By first face characteristic information and associated each second face characteristic information of multiple first user equipment informations Be compared, obtain associated each second face characteristic information of multiple first user equipment informations with it is described the first The similarity of face characteristic information；

Highest second face characteristic information of similarity is selected, as face recognition result.

8. according to any device of claim 5-7, which is characterized in that the second user facility information include it is following it One or any combination: body-worn information, the sequence number of mobile terminal, the MAC address of mobile terminal, movement Number, the client identification module SIM card sequence number of terminal；The content and described second that first user equipment information includes are used The content that family facility information includes is identical.