KR20170050980A - Face recognition system and method of adapting to environment - Google Patents

Face recognition system and method of adapting to environment Download PDF

Info

Publication number
KR20170050980A
KR20170050980A KR1020150153205A KR20150153205A KR20170050980A KR 20170050980 A KR20170050980 A KR 20170050980A KR 1020150153205 A KR1020150153205 A KR 1020150153205A KR 20150153205 A KR20150153205 A KR 20150153205A KR 20170050980 A KR20170050980 A KR 20170050980A
Authority
KR
South Korea
Prior art keywords
authentication
time
face
update
image
Prior art date
Application number
KR1020150153205A
Other languages
Korean (ko)
Other versions
KR101977887B1 (en
Inventor
정규택
Original Assignee
주식회사 파이브지티
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Family has litigation
First worldwide family litigation filed litigation Critical https://patents.darts-ip.com/?family=58741642&utm_source=google_patent&utm_medium=platform_link&utm_campaign=public_patent_search&patent=KR20170050980(A) "Global patent litigation dataset” by Darts-ip is licensed under a Creative Commons Attribution 4.0 International License.
Application filed by 주식회사 파이브지티 filed Critical 주식회사 파이브지티
Priority to KR1020150153205A priority Critical patent/KR101977887B1/en
Publication of KR20170050980A publication Critical patent/KR20170050980A/en
Application granted granted Critical
Publication of KR101977887B1 publication Critical patent/KR101977887B1/en

Links

Images

Classifications

    • G06K9/00221—
    • G06K9/00255—
    • G06K9/00261—
    • G06K9/62—

Landscapes

  • Collating Specific Patterns (AREA)

Abstract

More particularly, the present invention relates to an environmentally adaptive face recognition system and a method thereof, and more particularly, to a face recognition system in which authentication is likely to succeed according to a time when a user attempts authentication, The present invention relates to an environmentally adaptive face recognition system and a method thereof, in which a DB is automatically updated with facial images actually used for user authentication.
According to another aspect of the present invention, there is provided an authentication system, comprising: an authentication DB for classifying and storing a plurality of facial images in a plurality of time zones and storing information for updating facial images in each time period during which the user is authenticated; The face image stored in the authentication DB and the updated DB for storing the photographed image and the time taken to complete the authentication as the face image for updating and the authentication required time for updating the authentication DB, Since the DB management unit is provided for sorting the stored update face images in order of the required authentication time, the authentication DB is automatically updated to the face images that are actually used for the authentication of the user and can be expected to have a high authentication rate Therefore, the authentication rate and the face recognition speed in the face recognition process can be improved.

Description

[0001] The present invention relates to a face recognition system and a method of adapting to environment,

More particularly, the present invention relates to an environmentally adaptive face recognition system and a method thereof, and more particularly, to a face recognition system in which authentication is likely to succeed according to a time when a user attempts authentication, The present invention relates to an environmentally adaptive face recognition system and a method thereof, in which a DB is automatically updated with facial images actually used for user authentication.

The conventional face recognition system has a problem in that the authentication rate is lowered due to the difference between the surrounding environment at the time of initial registration of a face image per user and the surrounding environment at the time of actual authentication.

For example, in the conventional face recognition system, a face image photographed a plurality of times at the same time is recorded in a memory to register a face image for each user, and when a face image for actual authentication is inputted, When the face image is photographed by the user and the face image is photographed at a time when the face image is actually photographed, the change in the surrounding environment due to the difference in the time zone can not be reflected, so that the authentication rate is low .

That is, various variables such as the brightness and darkness of the user's surroundings and the difference in brightness of the user's face may be generated according to the time when the face image for actual authentication is captured. As a result, the authentication of the user fails or the authentication processing time increases do.

In addition, in the conventional face recognition system, when a face image for actual authentication is inputted, the face recognition speed in the authentication process is lowered because the face image is compared with the face images stored in the memory without any prior priority.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and it is an object of the present invention to provide an environmentally adaptive face recognition system and its method capable of improving the authentication rate and the face recognition speed in the face recognition process.

It is another object of the present invention to provide an environmentally adaptive face recognition system and a method thereof that can be compared with a captured image from a face image having a high possibility of authentication at a time when a user attempts authentication.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood by those skilled in the art from the following description.

According to another aspect of the present invention, there is provided a method for authenticating a face image, the method comprising: classifying a plurality of face images for face recognition into a plurality of time zones and storing information for updating the pre- DB; The time taken to complete the authentication and the photographed image photographed from the camera and matched with the face image of the authentication DB to complete the authentication is stored as the update face image for updating the authentication DB and the authentication required time Update DB; And a DB manager for replacing the face images stored in the authentication DB with update face images stored in the update DB and arranging them in the order of quickest authentication time when the preset condition is satisfied or the update cycle comes The present invention provides an environmentally adaptable face recognition system including the face recognition system.

In a preferred embodiment, authentication is performed by comparing the photographed image photographed by the camera with a plurality of face images stored in the authentication DB, wherein authentication is first performed by comparing the face images of the time zone corresponding to the input time of the photographed image And an authenticating unit for authenticating the authenticated user.

In a preferred embodiment, when the authentication of the face image in the time zone corresponding to the input time of the photographed image fails, the authentication unit compares the captured image with the photographed image in the order of the face images having the fastest authentication time.

In a preferred embodiment, a plurality of time zones, which are the same as the authentication DB, are set in the update DB, and the update face image and the authentication time are stored for each time zone in which the user has been authenticated.

In a preferred embodiment, the DB management unit compares the authentication required time of the update face image previously stored in the update DB with the authentication required time of the photographed image in which the user authentication is completed, This updates to a faster video.

In a preferred embodiment, the authentication DB stores the authentication minimum time, the authentication maximum time, the authentication average time, the number of authentication attempts, the number of authentication successes, and the authentication success rate for each time the user is authenticated, And stores it as information.

In a preferred embodiment, the DB management unit determines that the update condition of the authentication DB is satisfied when the average value of the authentication average times in all the time zones stored in the authentication DB is greater than a predetermined update reference time.

In a preferred embodiment, the DB management unit determines that the updating condition of the authentication DB is satisfied when the authentication success rate of all the time zones stored in the authentication DB is smaller than the predetermined success rate of update criteria.

In a preferred embodiment of the present invention, the DB management unit determines a time when the user's authentication is completed the fastest in each time zone as an authentication minimum time, a time when the user's authentication is delayed and completed late as an authentication maximum time, an average value of the authentication minimum time and the authentication maximum time And stores the average time in the authentication DB.

In a preferred embodiment, the DB management unit determines the number of authentication attempts, the number of times authentication of the face image stored in the corresponding time zone is successful, the authentication success count, the authentication attempt count, And stores the ratio of success times in the authentication DB as the authentication success rate.

According to another aspect of the present invention, there is provided a method for a face recognition system, comprising: (1) constructing an authentication DB in which a plurality of face images for face recognition are classified and stored in a plurality of time zones; (2) the face recognition system performs authentication by comparing the photographed image photographed by the camera with face images stored in the authentication DB; (3) The face recognition system stores information for updating the previously stored face images in the authentication DB for each time period during which the user is authenticated, matches the face image of the authentication DB photographed from the camera, Constructing an update DB storing the completed photographed image and the time taken to complete the authentication as the update face image and the authentication required time; (4) If the predetermined condition is satisfied or the update cycle comes, the face recognition system replaces the face images stored in the authentication DB with the update face images stored in the update DB, ; And (5) the face recognition system compares the photographed image photographed by the camera with a plurality of face images stored in the authentication DB to perform authentication, and firstly compares the face images of the time zone corresponding to the input time of the photographed image And performing authentication based on the face recognition result.

In a preferred embodiment of the present invention, in the step (5), the face recognition system first compares face images in a time zone corresponding to the input time of the captured image to attempt authentication, The authentication is performed by comparing the captured images with the facial images in the order of the fastest authentication time.

In a preferred embodiment, in the step (3), the face recognition system sets a plurality of time zones, which are the same as the authentication DB, in the update DB, and updates face images for authentication and authentication The time required is stored.

In a preferred embodiment of the present invention, in the step (3), the face recognition system may further include an authentication step of acquiring an authentication time of the update face image previously stored in the update DB, The time required for authentication is compared with the time required for authentication, and the image is updated to a faster image.

In a preferred embodiment, in the step (3), the face recognition system calculates the minimum authentication time, the authentication maximum time, the authentication average time, the authentication attempt count, the authentication success count, and the authentication success rate, And stores it in the authentication DB as information for updating the previously stored face images.

In a preferred embodiment, in the step (4), when the average value of the authentication average times in all the time periods stored in the authentication DB is larger than a predetermined update reference time, It is determined that the update condition is satisfied.

In a preferred embodiment, in the step (4), when the authentication success rate of all the time zones stored in the authentication DB is smaller than the predetermined success rate of update criteria, the face recognition system may update the authentication DB It is judged to be satisfied.

In a preferred embodiment of the present invention, in the step (3), the face recognition system calculates a time at which the user's authentication is completed most quickly for each time period as an authentication minimum time, a time at which a user's authentication is delayed and completed late as an authentication maximum time, And stores the average value of the time and the maximum authentication time in the authentication DB as the authentication average time.

In a preferred embodiment, in the step (3), the face recognition system determines the total number of times the user authentication is attempted in the corresponding time zone for each time zone, the number of authentication attempts, and the number of times authentication of the face image stored in the corresponding time zone is successful The ratio between the number of success times, the number of authentication attempts, and the number of successes of authentication is stored in the authentication DB as the authentication success rate.

According to the present invention, the present invention provides an authentication DB for classifying and storing a plurality of facial images in a plurality of time zones and storing information for updating the facial images for each time period during which the user is authenticated, An updating DB for storing the photographed image for which the authentication is completed and the time taken to complete the authentication as the updating face image for updating the authentication DB and the authentication required time and the face images stored in the authentication DB Since the DB management unit for replacing the face images for updating with the face images for updating stored in the update DB and arranging the authentication face images in the order of the fastest authentication time is provided, the authentication DB can be used as a face image It is possible to improve the authentication rate and the face recognition speed in the face recognition process.

In particular, when the user attempts to authenticate after the authentication DB is replaced with the photographed image actually used for authentication, the brightness of the user's surrounding environment at the time of the authentication attempt and the brightness Face image is first used for face recognition, so that user authentication can be completed more quickly and accurately.

In addition, even if the authentication of the face image in the time when the user tries to authenticate fails, the present invention is used for facial recognition from a face image having a high authentication time, which can expect a high authentication rate, .

1 is a view for explaining an environmentally adaptive face recognition system according to an embodiment of the present invention;
2 is a view for explaining an update DB of the environmentally adaptive face recognition system;
3 is a view for explaining an authentication DB of an environmentally adaptive face recognition system;
4 is a diagram for explaining an implementation example of an authentication DB.
5 is a view for explaining an environmentally adaptive face recognition method according to an embodiment of the present invention.
6 is a diagram for explaining an authentication process after replacing the authentication DB;

It should be understood that the specific details of the invention are set forth in the following description to provide a more thorough understanding of the present invention and that the present invention may be readily practiced without these specific details, It will be clear to those who have knowledge.

Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to FIGS. 1 to 6, and a description will be given centering on the parts necessary for understanding the operation and operation according to the present invention.

FIG. 1 is a view for explaining an environmentally adaptive face recognition system according to an embodiment of the present invention. FIG. 2 is a view for explaining an update DB of the environmentally adaptive face recognition system, FIG. 4 is a diagram for explaining an authentication DB of the face recognition system. FIG.

1 to 4, an environmentally adaptive face recognition system according to an embodiment of the present invention includes an authentication DB 110, an update DB 120, a DB management unit 130, and an authentication unit 140 .

The authentication DB 110 is a database in which the face images for user authentication and the information for updating the face images are classified and stored according to predetermined time zones. The update face images for updating may be provided as a database in which the face images are sorted and classified by time zone.

Here, the authentication DB 110 and the update DB 120 may store the facial image itself, or may extract and store only characteristic information used for facial recognition, and may also extract and store the facial image itself It is possible to store all of the feature information. It should be noted that the term "face image" used below is a concept including both. The authentication DB 110 and the update DB 120 may be provided as a plurality of databases separated from each other. In one database, the function of the authentication DB 110 and the function of the update DB 120 And the like.

The DB management unit 130 manages the authentication DB 110 and the update DB 120. The authentication DB 140 stores the authentication DB 110 and the update DB 120 in the authentication unit 140, 110), and the captured image including the user's face, for example, the captured image actually photographed from the camera and used for authentication is stored in the update DB 120 as the update face image Respectively.

The authentication unit 140 is for performing authentication of a user. The authentication unit 140 compares a photographed image photographed by a camera with a plurality of face images stored in the authentication DB 110 to recognize a face of a user to be authenticated Respectively.

Particularly, the authentication unit 140 can compare the facial images of the time zone corresponding to the input time of the captured images from the camera, among the facial images stored in the authentication DB 110 for each time zone, have.

Hereinafter, the detailed configuration of the environmentally adaptive face recognition system according to an embodiment of the present invention will be described in the order of the update DB 120 to the authentication DB 110. [

First, as shown in FIG. 2, a plurality of time zones are set in the update DB 120 according to an embodiment of the present invention, and an update face image and an authentication time are stored for each time zone.

Here, the plurality of time zones set in the update DB 120 may be divided into 2 to N intervals, for example, 24 hours, and the user may arbitrarily adjust the time zones. Also, the same time zone may be set in the authentication DB 110 as well.

The time required for authentication stored in the update DB 120 is a time required for the photographed image to be used for face recognition and to complete the authentication. The DB management unit 130 is used for face recognition for each time zone And stores the photographed image having the earliest time required for authentication among the photographed images that have been authenticated as an update face image.

For this purpose, the DB management unit 130 newly authenticates the user at the same time as the authentication time required for the update face image stored in the update DB 120 every time the user's authentication is completed and the captured image is input It is possible to store an image having a faster authentication time compared with the authentication time required for the captured image.

Also, when the predetermined condition is satisfied or the update cycle comes, the DB management unit 130 replaces the face images stored in the authentication DB 110 with the update face images stored in the update DB 120 . At this time, the authentication DB 110 stores information about authentication time and time zone for each update face image.

That is, since the update face images stored in the update DB 120 are actually used for authentication and are classified into time zones according to the authentication time, the user-specific face images initially registered in the authentication DB 110 are updated It is possible to expect an improved authentication rate and an authentication speed when the authentication unit 140 performs authentication.

Referring to FIG. 3, facial images for user's face recognition and information for updating facial images are classified into a plurality of time zones and stored in the authentication DB 110 of the present invention.

However, the authentication DB 110 shown in FIG. 3 is not initially registered, but is replaced at least once by the DB management unit 130 with the update face images stored in the update DB 120, It should be noted that the information for updating the images is accumulated.

That is, a plurality of time zones set in the authentication DB 110 are the same as a plurality of time zones set in the update DB 120, and the face images 1 to 4 are substantially stored in the update DB 120 The update face image 1 to the update face image 4 are updated. In addition, authentication required time for each update face image is stored together.

In addition, the DB management unit 130 arranges the face images of the authentication DB 110 in the order of the fastest authentication time, and when the authentication of the user is completed by the authentication unit 140, The authentication maximum time, the authentication average time, the number of authentication attempts, the number of authentication successes, and the authentication success rate are stored in the authentication DB 110. Here,

First, the authentication minimum time is stored as the time when the authentication of the user is completed fastest by the authentication unit 140 for each time zone. On the other hand, the time at which the user's authentication is completed latest at each time slot is stored as the authentication maximum time, The authentication minimum time and the average value of the authentication maximum time are stored as the authentication average time. For reference, the minimum authentication time is set at the same time as the authentication required time at first, and may be changed when the authentication of the user is completed earlier than the previously stored authentication required time.

Also, the number of authentication attempts is stored as the total number of times user authentication is attempted for each time slot, the number of authentication successes is stored as the number of times authentication is completed for each time slot, and the authentication success rate is stored as a ratio between the authentication attempt count and the authentication success count . Meanwhile, the number of authentication attempts, the number of authentication successes, and the authentication success rate may be integrally calculated and stored for the authentication attempts and the success times for all the time zones.

For example, the DB management unit 130 increases the number of authentication attempts in the time zone in which the user authentication is attempted by +1 and stores the increase in the number of authentication attempts in the time zone in which the user authentication is attempted. . However, when the authentication is not successful with the face image of the corresponding time zone, the number of authentication attempts is increased without changing the number of authentication attempts.

In addition, the information for updating the facial images stored in the authentication DB 110 is used as information for determining whether the DB manager 130 has arrived at a predetermined condition or a renewal cycle.

That is, if the average value of the authentication average times in all the time zones stored in the authentication DB 110 is greater than the predetermined update reference time, or if the average value of the authentication success rates in all the time zones is predefined It is determined that the update condition of the authentication DB 110 is satisfied or the update cycle has arrived and the face images of all the time periods stored in the authentication DB 110 are stored in the update DB 120. [ The face images for updating can be replaced with the face images for updating stored in the memory.

Meanwhile, if the authentication average time in the specific time zone is greater than the update reference time defined in advance, or if the authentication success rate in the specific time zone is smaller than the predetermined update rule success rate, May be replaced with the update face image classified and stored in a specific time zone in the update DB 120. [

For example, in FIG. 3, the authentication success rate in the "a" time zone is stored at 84%, which is relatively low compared to the authentication success rate in the other time zones. At this time, assuming that the previously defined update criterion success rate is 90%, the DB management unit 130 replaces the face image of the user A in the time zone "a" with the face image for update stored in the update DB 120 do.

In addition, the DB management unit 130 determines whether or not the face image is replaced for each user registered in the authentication DB 110, and replaces it with the update face images stored for each user in the update DB 120 You may.

The DB management unit 130 compares the time required for authentication of the face image stored in the authentication DB 110 with the time required for authentication of the update face image in the corresponding time zone, If the time required for authentication of the previously stored facial image is earlier, it may not be replaced with the update facial image.

In addition, the DB management unit 130 may replace the face images of the authentication DB 110 with the face images for update of the update DB 120, Information for updating the images, that is, the authentication minimum time, the authentication maximum time, the authentication average time, the authentication attempt count, the authentication success count, and the authentication success rate are reset and the update face images of the update DB 120 are reset.

In addition, the DB management unit 130 may determine that the average value of the authentication average times in all the time zones is significantly larger than the update reference time defined in advance, or the average value of the authentication success rates in all the time zones If the defined update criterion is significantly less than the success rate, the user may be asked to re-register the facial images for a particular user.

FIG. 3 exemplarily shows information for updating the face images and the face images of the user A, but the authentication DB 110 substantially updates the face images of all users with the update of the face images Are stored.

Referring to FIG. 4, in the authentication DB 110 of the present invention, substantially all the user's facial images are arranged in the order of fastest authentication time, and each facial image is classified according to the time zone in which the authentication is actually performed have. For all users, the minimum authentication time, the maximum authentication time, the average authentication time, the number of authentication attempts, the number of authentication successes, and the authentication success rate are accumulated and stored for each time slot.

Accordingly, the authentication unit 10 can use the authentication DB 110 to compare a face image having a high possibility of authentication, that is, a face image having a short authentication required time, with a face of a user to be authenticated.

For example, the authentication unit 140 may compare the facial image of the time zone corresponding to the input time of the photographed image photographed by the camera, among the facial images classified by time in the authentication DB 110, with the photographed image first .

That is, when the authentication of the user is requested in the " a " time zone or the captured image of the user who intends to authenticate is input, the authentication unit 140 reads the face image stored in the authentication DB 110, The face image 1 of the user B, the face image 1 of the user A, and the face image 1 of the user D, which are facial images of the user B, can be compared with the captured image to recognize the user's face.

When the authentication of the face image in the time zone corresponding to the input time of the photographed image fails, the authentication unit 140 performs authentication by comparing the photographed image with the photographed image in the order of the face images having the fastest authentication time, Or authentication failure can be determined. At this time, the process of comparing the face images of all the time periods excluding the face images of the " A " time zone with the shot images in the order of the fastest authentication time can be performed.

In addition, the authentication unit 140 outputs an authentication result after performing the authentication, and a follow-up process such as identity verification, commute management, opening a door can be performed according to the authentication result.

Accordingly, the environmentally adaptive face recognition system according to an embodiment of the present invention stores the photographed image photographed at the time when the actual authentication is attempted in the authentication DB 110 and uses it as a face image for authentication, The brightness and darkness of the user's surroundings and the contrast difference of the user's face are reflected upon recognition, thereby minimizing the authentication failure and improving the authentication speed and authentication rate.

Also, even if the authentication of the facial image at the time when the user tries to authenticate fails, the authentication unit 140 uses the facial image for the face recognition in order of the fastest authentication time required to expect a high authentication rate, User authentication can be completed.

FIG. 5 is a view for explaining an environmentally adaptive face recognition method according to an embodiment of the present invention, and FIG. 6 is a view for explaining an authentication process after replacing the authentication DB.

5 and 6, an environmentally adaptive face recognition method performed in the environmentally adaptive face recognition system according to an embodiment of the present invention will be described.

However, since the functions performed in the environmentally adaptive face recognition method shown in Figs. 5 and 6 are all performed in the environmentally adaptive face recognition system described with reference to Figs. 1 to 4, All the functions described with reference to FIG. 4 are performed in the environmentally adaptive face recognition method according to the preferred embodiment of the present invention, and all the functions described with reference to FIGS. 5 and 6 are performed according to the environmental adaptation Type face recognition system.

First, an authentication DB storing a plurality of face images for each user is constructed (S110).

At this time, in the state where the authentication DB or DB management unit is set to the user image registration mode, the face images of the face images of the user's face are initially registered, the initially registered face images are classified and stored in a plurality of time zones, Or may not have a separate classification criterion.

Next, the authentication unit compares the photographed image photographed by the camera with the facial images stored in the authentication DB to perform authentication (S120).

At this time, the authentication unit compares the images themselves, compares the characteristic information of each image, or compares the image itself and characteristic information, and compares the captured image with the face images stored in the authentication DB to perform authentication do.

Next, when the authentication is successful, the DB management unit constructs an update DB storing the image for update for updating the authentication DB (S130).

At this time, the DB management unit sets a plurality of time zones in the update DB, and stores the photographed image, which has been authenticated each time the user's authentication is completed, as the update face image in the time zone corresponding to the authentication time. That is, a plurality of update face images are classified in the predetermined time zone and stored in the update DB.

Preferably, the DB management unit compares the authentication required time of the photographed image with the authentication required time of the update face image in the same time zone already stored in the update DB every time the user authentication is completed, . Here, the authentication required time means the time taken for the update image or the photographed image to complete the authentication.

In addition, the DB management unit stores information for updating the authentication DB together with the face image for update in the authentication DB for each time zone. For example, the DB management unit updates at least one of the authentication minimum time, the authentication maximum time, the authentication average time, the authentication attempt count, the authentication success count, and the authentication success rate in the authentication DB in the time zone in which the authentication is performed.

At this time, the DB management unit stores the time when the user's authentication is completed fastest as the authentication minimum time, stores the time when the user's authentication is delayed and completed late as the authentication maximum time, calculates the average value of the authentication minimum time and the authentication maximum time The average authentication time is stored.

Also, the DB management unit increases the number of authentication attempts in the time zone in which the user authentication is attempted by +1 and stores the increased number of authentication attempts by +1. If the authentication succeeds in the face image in the corresponding time zone, The ratio between the number of times of authentication and the number of authentication successes is calculated and stored as the authentication success rate.

Meanwhile, when the authentication is performed in a specific time zone but the authentication is not performed with the face image of the corresponding time zone but the authentication is performed with the face image of another time zone, the DB management unit only increases the number of authentication attempts in the time zone in which the user authentication is attempted, The number of successes may be left unchanged. In this case, the authentication success rate in a specific time period is reduced.

Next, the face images of the authentication DB are replaced by the DB management unit (S140).

At this time, the DB management unit replaces the face image of the authentication DB with the update face images corresponding to the respective time zones of the update DB, when the predetermined condition is satisfied or the update cycle comes (S141) And a step of aligning the replaced face images in the order of the fastest authentication time (S142).

To this end, when the average value of the authentication average times in all the time zones stored in the authentication DB is larger than the update reference time defined in advance, or when the authentication success rate is smaller than the predefined update criterion success rate, It can be determined that the update period of the DB has arrived. On the other hand, an arbitrary update period may be set by the user.

Also, in the updating process of the authentication DB, the DB manager may replace the face images having the authentication success rate lower than the update reference success rate with the face image for update, or the authentication average time is shorter than the update reference time The large face images are replaced with the update face images.

The DB management unit compares the authentication time required for the facial image stored in the authentication DB with the authentication time required for the update face image in the corresponding time zone, It may not be replaced with the update face image.

When the face images of the authentication DB are replaced with the face images for updating, the authentication minimum time, the authentication maximum time, the authentication average time, the authentication attempt count, the authentication success count and the authentication success rate stored in the authentication DB are reset do. Similarly, the face images for update of the update DB can also be reset.

Next, the authentication unit compares the photographed image photographed by the camera with face images stored in the authentication DB to perform authentication (S150).

At this time, the authentication unit compares the face image of the time zone corresponding to the input time of the photographed image with the photographed image to perform authentication (S151). If there is no matching image (S152) The authentication is performed by comparing the captured image with the images in step S153.

In addition, even when the facial image and the photographed image are compared in order of the required authentication time, if there is no matching image (S154), the authentication result is output as failure of the authentication failure (S155).

In addition, if there is an identical image in steps 152 and 154, the authentication unit outputs an authentication result indicating that the authentication is successful (S156). In addition, a follow-up process such as identity verification, commute management, door opening and the like can be performed according to the authentication result of the authentication unit.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the invention.

110: DB for authentication
120: DB for updating
130: DB manager
140:

Claims (19)

An authentication DB for classifying and storing a plurality of face images for face recognition in a plurality of time zones and storing information for updating the previously stored face images for each time period during which the user is authenticated;
The time taken to complete the authentication and the photographed image photographed from the camera and matched with the face image of the authentication DB to complete the authentication is stored as the update face image for updating the authentication DB and the authentication required time Update DB; And
And a DB management unit for replacing face images stored in the authentication DB with update face images stored in the update DB and arranging them in the order of the fastest authentication time when the predetermined condition is satisfied or the update cycle comes Based face recognition system.
The method according to claim 1,
An authentication unit for performing authentication by comparing an image photographed by a camera with a plurality of face images stored in the authentication DB and comparing the face images of the time zone corresponding to the input time of the photographed image first to perform authentication Wherein the environment-aware face recognition system comprises:
3. The method of claim 2,
Wherein when the authentication of the face image in the time zone corresponding to the input time of the photographed image fails, the authentication unit performs authentication by comparing the photographed image with the photographed image in the order of the face image having the fastest authentication time, system.
The method according to claim 1,
Wherein the update DB stores a plurality of time zones, which are the same as the authentication DB, and stores an update face image and an authentication time for each time zone in which the user has been authenticated.
5. The method of claim 4,
The DB management unit compares the authentication required time of the update face image previously stored in the update DB with the authentication required time of the photographed image in which the user authentication is completed for each time zone in which the user authentication is completed, Wherein the face recognition system comprises:
The method according to claim 1,
The authentication DB stores the authentication minimum time, the authentication maximum time, the authentication average time, the authentication attempt count, the authentication success count, and the authentication success rate as information for updating the pre-stored face images for each time period during which the user is authenticated And a face recognition system.
The method according to claim 6,
Wherein the DB management unit determines that the update condition of the authentication DB is satisfied when the average value of the authentication average times in all the time periods stored in the authentication DB is larger than a predetermined update reference time, Face Recognition System.
The method according to claim 6,
Wherein the DB management unit determines that the updating condition of the authentication DB is satisfied when the authentication success rate of all the time zones stored in the authentication DB is smaller than the predetermined success rate of the update criterion success rate .
The method according to claim 6,
The DB management unit stores the authentication time of the user in each time zone as the authentication minimum time, the authentication completion time of the user, the authentication maximum time, the authentication minimum time, and the authentication maximum time as the authentication average time. DB in the face recognition system.
The method according to claim 6,
The DB management unit calculates a ratio of the number of authentication attempts, the number of times authentication of the face image stored in the corresponding time zone is successful to the number of authentication success times, the ratio of the number of authentication attempts to the number of authentication success times, And stores it in the authentication DB at a success rate.
(1) building a DB for authentication in which a plurality of face images for face recognition are classified and stored in a plurality of time zones;
(2) the face recognition system performs authentication by comparing the photographed image photographed by the camera with face images stored in the authentication DB;
(3) The face recognition system stores information for updating the previously stored face images in the authentication DB for each time period during which the user is authenticated, matches the face image of the authentication DB photographed from the camera, Constructing an update DB storing the completed photographed image and the time taken to complete the authentication as the update face image and the authentication required time;
(4) If the predetermined condition is satisfied or the update cycle comes, the face recognition system replaces the face images stored in the authentication DB with the update face images stored in the update DB, ; And
(5) The face recognition system compares the photographed image photographed by the camera with a plurality of facial images stored in the authentication DB to perform authentication, and first compares facial images in a time zone corresponding to the input time of the photographed image And performing authentication based on the face recognition result.
12. The method of claim 11,
In the step (5), the face recognition system first compares face images in a time zone corresponding to the input time of the captured image to try authentication, and when authentication fails, And the authentication is performed in comparison with the captured image in the order of the quick face images.
12. The method of claim 11,
In the step (3), the face recognition system sets a plurality of time zones, which are the same as the authentication DB, in the update DB, and stores the update face image and the authentication required time for each time zone in which the user has been authenticated A method for recognizing an environmentally adaptive face.
14. The method of claim 13,
In the step (3), the face recognition system compares the authentication required time of the update face image previously stored in the update DB with the authentication required time of the user image And updating the image with a faster authentication time.
12. The method of claim 11,
In the step (3), the face recognition system calculates the minimum authentication time, the authentication maximum time, the authentication average time, the authentication attempt count, the authentication success count, and the authentication success rate for each time period during which the user is authenticated And storing the information in the authentication DB.
16. The method of claim 15,
In the step (4), the face recognition system satisfies the updating condition of the authentication DB when the average value of the authentication average times in all the time periods stored in the authentication DB is larger than the predetermined update reference time Wherein the face recognition method comprises the steps of:
16. The method of claim 15,
In the step (4), the face recognition system determines that the update condition of the authentication DB is satisfied when the authentication success rate of all the time periods stored in the authentication DB is smaller than the predetermined success rate A method for recognizing an environmentally adaptive face.
16. The method of claim 15,
In the step (3), the face recognition system determines the time at which the user's authentication is completed the fastest in each time zone as the authentication minimum time, the time when the user's authentication is delayed and completed late as the authentication maximum time, And stores the average value as the authentication average time in the authentication DB.
16. The method of claim 15,
In the step (3), the face recognition system calculates the total number of times the user authentication is attempted in the corresponding time zone for each time period as the number of authentication attempts, the number of successful authentication of the face image stored in the corresponding time zone, And the authentication success rate is stored in the authentication DB as the authentication success rate.
KR1020150153205A 2015-11-02 2015-11-02 Face recognition system and method of adapting to environment KR101977887B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020150153205A KR101977887B1 (en) 2015-11-02 2015-11-02 Face recognition system and method of adapting to environment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020150153205A KR101977887B1 (en) 2015-11-02 2015-11-02 Face recognition system and method of adapting to environment

Publications (2)

Publication Number Publication Date
KR20170050980A true KR20170050980A (en) 2017-05-11
KR101977887B1 KR101977887B1 (en) 2019-05-13

Family

ID=58741642

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020150153205A KR101977887B1 (en) 2015-11-02 2015-11-02 Face recognition system and method of adapting to environment

Country Status (1)

Country Link
KR (1) KR101977887B1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107507216A (en) * 2017-08-17 2017-12-22 北京觅己科技有限公司 The replacement method of regional area, device and storage medium in image
KR20220032244A (en) * 2020-09-07 2022-03-15 (주)에이쓰리시큐리티 Disinfection system using thermal image and visible image

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010237756A (en) * 2009-03-30 2010-10-21 Fujifilm Corp Method and device for authenticating face
JP2012208610A (en) * 2011-03-29 2012-10-25 Secom Co Ltd Face image authentication apparatus
KR101297736B1 (en) * 2013-05-07 2013-08-23 주식회사 파이브지티 Face recognition method and system

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010237756A (en) * 2009-03-30 2010-10-21 Fujifilm Corp Method and device for authenticating face
JP2012208610A (en) * 2011-03-29 2012-10-25 Secom Co Ltd Face image authentication apparatus
KR101297736B1 (en) * 2013-05-07 2013-08-23 주식회사 파이브지티 Face recognition method and system

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107507216A (en) * 2017-08-17 2017-12-22 北京觅己科技有限公司 The replacement method of regional area, device and storage medium in image
CN107507216B (en) * 2017-08-17 2020-06-09 北京觅己科技有限公司 Method and device for replacing local area in image and storage medium
KR20220032244A (en) * 2020-09-07 2022-03-15 (주)에이쓰리시큐리티 Disinfection system using thermal image and visible image

Also Published As

Publication number Publication date
KR101977887B1 (en) 2019-05-13

Similar Documents

Publication Publication Date Title
US20200175263A1 (en) Biometric identification and verification
US10380499B2 (en) Machine-learning system for optimising the performance of a biometric system
US10503992B2 (en) Process for updating templates used in facial recognition
US20190213394A1 (en) Multiple enrollments in facial recognition
EP3182316B1 (en) Fingerprint authentication method and system, and terminal supporting fingerprint authentication
US20180004924A1 (en) Systems and methods for detecting biometric template aging
JP2017091520A (en) Methods and apparatuses for adaptively updating enrollment database for user authentication
US20110138187A1 (en) System and method of biometric authentication using multiple kinds of templates
JP2004157602A (en) Apparatus and method for person recognition and passage controller
CN108734092B (en) Person authentication device
US20150205995A1 (en) Personal recognition apparatus that performs personal recognition using face detecting function, personal recognition method, and storage medium
CN110612530A (en) Method for selecting a frame for use in face processing
KR101977887B1 (en) Face recognition system and method of adapting to environment
KR20180102574A (en) How to update biometric criteria data
KR102040482B1 (en) Fingerprint template management method and authentication apparatus using fingerprint template management method thereof
JP2018128736A (en) Face authentication system, face authentication method and face authentication program
US9648112B2 (en) Electronic device and method for setting network model
KR20070023022A (en) Method For Updating A Biological Information
JP2018169872A (en) Face authentication system, face authentication method, and face authentication program
CN114140822A (en) Pedestrian re-identification method and device
WO2020261906A1 (en) Image processing device and method for same, image input device, image processing system, and program
JPWO2019207649A1 (en) Authentication device
KR102274370B1 (en) Facial recognition device and facial recognition control method
JP2010003010A (en) Face authentication device and face authentication method
CN112818784B (en) Control method and device of access control equipment and storage medium

Legal Events

Date Code Title Description
A201 Request for examination
E902 Notification of reason for refusal
E701 Decision to grant or registration of patent right
GRNT Written decision to grant