JP6432634B2 - Authentication device, authentication method, and program - Google Patents

Description

  The present invention relates to an authentication apparatus and an authentication method for performing authentication processing using image data, and further relates to a program for realizing these.

  In recent years, in an amusement facility, a public facility, a store, and the like, use of a face authentication system that automatically identifies a person from an image captured by a camera is increasing in order to increase security. The face authentication system identifies a person by comparing a face image registered in advance as a master and a photographed image based on feature points.

  By the way, in the conventional face authentication system, the person is identified by comparing the feature points between the master and the image. Therefore, if the face angle in the photographed image is slightly different, the matching may not be possible. is there.

  For this reason, for example, Patent Document 1 discloses a face authentication system that registers a plurality of masters having different face angles with respect to one person. The face authentication system disclosed in Patent Document 1 determines the face angle of a captured face image in face authentication, and performs face authentication using the master of the determined angle and the captured face image. Further, the face authentication system disclosed in Patent Document 1 selects an image with the highest recognition rate from an image used as a master and a captured image, and an image in which the selected image is captured. If so, update the master.

  According to the face authentication system disclosed in Patent Document 1, it is considered that the problem that the face angle is different and cannot be collated can be solved.

Japanese Patent No. 5785667

  However, in the face authentication system disclosed in Patent Document 1, in order to register an image with a different face angle as a master, it is necessary to perform master registration for the required number of masters for each set angle. . For this reason, the face authentication system disclosed in Patent Document 1 has a problem that it takes too much time and burden on the system administrator and the person to be authenticated.

  An example of an object of the present invention is to provide an authentication device, an authentication method, and a program capable of solving the above-described problem and increasing the number of master variations and improving authentication accuracy without requiring manual processing. There is to do.

In order to achieve the above object, an authentication apparatus according to one aspect of the present invention provides:
An image acquisition unit that acquires a plurality of images that are continuous along the time series of the subject from the imaging device;
An authentication processing unit that determines whether or not each of the plurality of images matches a registered image registered in advance in the order that the plurality of images are captured by the imaging device, and performs authentication based on the determination result;
An image registration unit that registers, in association with the registered image, an image that has not been determined to match when the number of times that it has been determined to match is equal to or greater than a set number;
It is characterized by having.

In order to achieve the above object, an authentication method according to one aspect of the present invention includes:
(A) acquiring a plurality of continuous images along the time series of the subject from the imaging device;
(B) determining whether each of the plurality of images matches a registered image registered in advance in the order taken by the imaging device, and performing authentication based on the determination result;
(C) when the number of times of continuous determination as matching is equal to or greater than a set number, registering an image that has not been determined to match in association with the registered image;
It is characterized by having.

Furthermore, in order to achieve the above object, a program according to one aspect of the present invention is provided.
On the computer,
(A) acquiring a plurality of continuous images along the time series of the subject from the imaging device;
(B) determining whether each of the plurality of images matches a registered image registered in advance in the order taken by the imaging device, and performing authentication based on the determination result;
(C) when the number of times of continuous determination as matching is equal to or greater than a set number, registering an image that has not been determined to match in association with the registered image;
Is executed.

  As described above, according to the present invention, it is possible to increase the number of master variations and improve the authentication accuracy without requiring manual processing.

FIG. 1 is a block diagram showing a schematic configuration of an authentication apparatus according to an embodiment of the present invention. FIG. 2 is a block diagram specifically showing the configuration of the authentication device according to the embodiment of the present invention. FIG. 3 is a flowchart showing the operation of the authentication apparatus in the embodiment of the present invention. FIG. 4 is a diagram showing an example of a continuous image acquired in the embodiment of the present invention. FIG. 5 is a diagram showing an example of an image stored in the storage unit in the embodiment of the present invention. FIG. 6 is a block diagram illustrating an example of a computer that implements the authentication apparatus according to the embodiment of the present invention.

(Embodiment)
Hereinafter, an authentication device, an authentication method, and a program according to an embodiment of the present invention will be described with reference to FIGS.

[Device configuration]
First, a schematic configuration of the authentication device according to the present embodiment will be described. FIG. 1 is a block diagram showing a schematic configuration of an authentication apparatus according to an embodiment of the present invention.

  An authentication device 10 according to the present embodiment shown in FIG. 1 is a device that performs authentication using an image captured by the imaging device 20. As shown in FIG. 1, the authentication device 10 includes an image acquisition unit 11, an authentication processing unit 12, and an image registration unit 13.

  The image acquisition unit 11 acquires a plurality of continuous images from the imaging device 20 along the time series of the subject. The authentication processing unit 12 determines whether or not each of the plurality of images matches a registered image (hereinafter also referred to as “master”) registered in advance in the order in which the plurality of images were captured by the imaging device 20. Based on authentication. The image registration unit 13 registers, in association with the registered image, an image that has not been determined to match when the number of times that it has been determined to match is equal to or greater than the set number.

  As described above, in the present embodiment, the authentication process is performed using continuously captured images, and when it is determined that the images match continuously, the images match in the series of images. An image that has not been determined (hereinafter referred to as “non-matching image”) is extracted, and this image is registered.

  That is, the non-matching image is an image that is continuous with an image determined to match (hereinafter referred to as “matching image”), and the subject is the same in both images. Since authentication is allowed for this subject, the non-matching image is also an image that should be one of the masters. Therefore, according to the present embodiment, it is possible to increase the number of master variations and improve the authentication accuracy without requiring manual processing.

  Next, the configuration of the authentication device 10 according to the present embodiment will be described more specifically with reference to FIG. FIG. 2 is a block diagram specifically showing the configuration of the authentication device according to the embodiment of the present invention. In the following, a case will be described in which the authentication device 10 is a device that performs person face authentication and the registered image (master) is a person face image.

  As shown in FIG. 2, in the present embodiment, the authentication device 10 includes a storage unit 14 and a master database 15 in addition to the image acquisition unit 11, the authentication processing unit 12, and the image registration unit 13 described above. ing.

  In the present embodiment, the image acquisition unit 11 acquires image data output by the imaging device 20 at set intervals, and stores the acquired image data in the storage unit 14 in time series, that is, in the order of shooting. To do.

  In this embodiment, the authentication processing unit 12 reads out image data from the storage unit 14 in the order stored in the storage unit 14 and collates with the master registered in the master database 15 for each read image data. Specifically, the authentication processing unit 12 first detects a human face image from the read image data, and further extracts features from the detected face image. Subsequently, the authentication processing unit 12 collates the extracted features with the features of each master registered in the master database 15.

  In addition, the authentication processing unit 12 compares and calculates parameters such as the position and size of the face, brightness and darkness, and moving distance for each image data. If the authentication processing unit 12 clearly cannot be used for authentication, the processing can be stopped.

  Then, in any of the read image data, the authentication processing unit 12 specifies the master having the matched feature when the feature extracted from the image data matches the feature of any master. Also, the authentication processing unit 12 determines that the person shown in the read image data is the master person who can be identified.

  In the present embodiment, the image registration unit 13 identifies image data whose image features match those of the master, and sets the number of image data continuous in time series among the identified image data. Determine if it is greater than or equal to the number.

  In addition, the image registration unit 13 determines that the image data that has not been determined to match the feature of the master (non-matching image) when the number of continuous image data is greater than or equal to the set number. ), The coincidence ratio of the image data determined to match the feature of the master with respect to the image (matching image) is calculated. Specifically, the image registration unit 13 calculates the similarity between the feature extracted from the image data of the non-matching image and the feature extracted from the image data of the matching image, and sets the calculated similarity as the matching rate. Can do.

  Furthermore, the image registration unit 13 determines whether an image of image data (non-matching image) that has not been determined to have the same image feature as the master feature satisfies the setting requirements for image quality. This is because, if the image quality of the non-matching image is equal to or higher than a certain standard, it is possible to avoid a situation where the authentication accuracy is lowered due to the image quality when the non-matching image is registered as a master.

  The setting requirements for image quality include the resolution of the image, the degree of contrast, the size of the face image, the degree of face-likeness, and the like, and the threshold value for each requirement is set as appropriate through experiments.

  Then, the image registration unit 13 associates the image data of the non-matching image with the already registered master when the calculated matching rate is equal to or higher than the threshold and the image quality of the non-matching image satisfies the setting requirement. This is registered in the master database 15 as a new master.

[Device operation]
Next, operation | movement of the authentication apparatus 10 in this Embodiment is demonstrated using FIG. FIG. 3 is a flowchart showing the operation of the authentication apparatus in the embodiment of the present invention. In the following description, FIGS. 1 and 2 are referred to as appropriate. Further, in the present embodiment, the authentication method is performed by operating the authentication device 10. Therefore, the description of the authentication method in the present embodiment is replaced with the following description of the operation of the authentication device 10.

  As shown in FIG. 3, first, the image acquisition unit 11 acquires image data output by the imaging device 20 at a set interval, and stores the acquired image data in the storage unit 14 in time series (step A1). ).

  Next, the authentication processing unit 12 reads the image data from the storage unit 14 in the order stored in the storage unit 14, and collates with the master registered in the master database 15 for each read image data (step A2). . In step A2, if the image feature of any image data matches the feature of the master, the authentication processing unit 12 determines that the person shown in the image data is the master person. .

  Next, the image registration unit 13 determines whether or not the authentication is successful in Step A2 (Step A3). If the result of determination in step A3 is that authentication has not succeeded, the image registration unit 13 ends the process.

  On the other hand, if the authentication is successful as a result of the determination in step A3, the image registration unit 13 identifies image data in which the image characteristics match the master characteristics, and among the identified image data, the time series. It is determined whether or not the number of image data continuous along the line is greater than or equal to the set number (step A4). As a result of the determination in step A4, if the number of image data continuous along the time series is not equal to or greater than the set number, the image registration unit 13 ends the process.

  On the other hand, as a result of the determination in step A4, when the number of image data continuous along the time series is equal to or larger than the set number, the image registration unit 13 calculates a coincidence rate with respect to the coincidence image in the non coincidence image ( Step A5).

  Next, the image registration unit 13 determines whether or not the matching rate calculated in step A5 is equal to or greater than a threshold value (step A6). If the result of determination in step A6 is that the matching rate is not equal to or greater than the threshold value, the image registration unit 13 ends the process.

  On the other hand, as a result of the determination in step A6, if the matching rate is equal to or greater than the threshold, the image registration unit 13 determines whether the non-matching image satisfies the setting requirements for image quality (step A7). As a result of the determination in step A7, when the setting requirements are not satisfied, the image registration unit 13 ends the process.

  On the other hand, as a result of the determination in step A7, if the setting requirement is satisfied, the image registration unit 13 associates the image data of the non-matching image with the already registered master, and uses this as a new master, the master database 15 (Step A8).

[Concrete example]
Next, a specific example of the operation of the authentication device 10 will be described with reference to FIGS. 4 and 5. FIG. 4 is a diagram showing an example of a continuous image acquired in the embodiment of the present invention. FIG. 5 is a diagram showing an example of an image stored in the storage unit in the embodiment of the present invention.

  As illustrated in FIG. 4, it is assumed that the imaging device 20 continuously captures a person A, and the image data of the face of the person A is continuously output in time series. The N pieces of image data output at this time are acquired by the image acquisition unit 11, and then stored as a new image 1, a new image 2,..., A new image N in the storage unit 14, as shown in FIG. Is done. N is an arbitrary natural number.

  The authentication processing unit 12 sequentially extracts new images stored in the storage unit 14, and collates the characteristics of the extracted new images with the characteristics of the master. As a result, in the image data (N-2), the image data (N-1), and the image data (N), it is assumed that the feature of the image matches the feature of the master of the person A. In this case, the authentication processing unit 12 authenticates the photographed person as the person A.

  Next, the image registration unit 13 associates the image data that matches the master of the person A with the master of the person A. Further, in this specific example, three or more pieces of image data (N-2) to image data (N) that are continuous in time series match the master of person A. Therefore, it is considered that the person shown in the image data (N-3) immediately before the image data (N-2) does not match in the collation but is likely to be the person A. The same applies to image data 1 to (N-4).

  Therefore, the image registration unit 13 calculates the coincidence rate of the image data 1 to (N-3) with the image data (N-2) determined to be coincident. Further, the image registration unit 13 determines whether the image quality satisfies the setting requirement for image data having a matching rate equal to or higher than a threshold value. In this specific example, the image data (N-3) has a matching rate equal to or higher than the threshold value, and the image quality satisfies the setting requirements. Therefore, the image registration unit 13 stores the image data (N-3) as the person A. Are registered in the master database 15 as the master of person A.

  In this way, in the image data having a high matching rate with the image data that has been successfully authenticated, the face of the person in the image data is near the boundary where it can be authenticated or not with respect to the master image facing the front. , Facing sideways, or changing facial expressions. In the present embodiment, since an image showing such a face is registered as a master, an improvement in authentication accuracy can be expected.

[Effects of the embodiment]
As described above, according to the present embodiment, in addition to the front face registered in advance as a master, the range of faces that can be authenticated can be gradually expanded. In other words, in the present embodiment, an image that may have been determined to be the original person but has not been used for collation determination is automatically picked up while operating the authentication device. Can be improved.

  Therefore, according to the present embodiment, re-registration of the master by the administrator and the person to be authenticated becomes unnecessary, and the management man-hour and convenience are improved. Furthermore, according to the present embodiment, it is possible to follow a person's secular change. Further, conventionally, there has been a problem that authentication cannot be performed successfully unless the face is close to the first registered face, a close angle, or the like, but this embodiment can solve such a problem.

  In the above-described example, the authentication device 10 in the present embodiment is used for face authentication. However, the authentication device 10 is not limited to this, and may be a system that extracts feature points from an image or video and performs matching. If applicable. Further, the present embodiment can be applied to a case where a plurality of imaging devices are provided and a thermosensor is used. According to these cases, further improvement in authentication accuracy can be expected. Further, the imaging device 10 may output a still image or may output a moving image. Further, the connection between the imaging device 20 and the authentication device 10 may be performed by either wired or wireless connection.

[program]
The program in the present embodiment may be a program that causes a computer to execute steps A1 to A8 shown in FIG. By installing and executing this program on a computer, the authentication device 10 and the authentication method in the present embodiment can be realized. In this case, a CPU (Central Processing Unit) of the computer functions as an image acquisition unit 11, an authentication processing unit 12, and an image registration unit 13, and performs processing.

  In the present embodiment, the storage unit 14 and the master database 15 are stored by storing data files constituting them in a storage device such as a hard disk provided in the computer, or the data files are stored. This can be realized by mounting the recording medium on a reading device connected to a computer.

  The program in the present embodiment may be executed by a computer system constructed by a plurality of computers. In this case, for example, each computer may function as any of the image acquisition unit 11, the authentication processing unit 12, and the image registration unit 13. The storage unit 14 and the master database 15 may be constructed on a computer different from the computer that executes the program in the present embodiment.

  Here, a computer that realizes the authentication device 10 by executing the program according to the present embodiment will be described with reference to FIG. FIG. 6 is a block diagram illustrating an example of a computer that implements the authentication apparatus according to the embodiment of the present invention.

  As shown in FIG. 6, the computer 110 includes a CPU 111, a main memory 112, a storage device 113, an input interface 114, a display controller 115, a data reader / writer 116, and a communication interface 117. These units are connected to each other via a bus 121 so that data communication is possible.

  The CPU 111 performs various calculations by developing the program (code) in the present embodiment stored in the storage device 113 in the main memory 112 and executing them in a predetermined order. The main memory 112 is typically a volatile storage device such as a DRAM (Dynamic Random Access Memory). Further, the program in the present embodiment is provided in a state of being stored in a computer-readable recording medium 120. Note that the program in the present embodiment may be distributed on the Internet connected via the communication interface 117.

  Specific examples of the storage device 113 include a hard disk drive and a semiconductor storage device such as a flash memory. The input interface 114 mediates data transmission between the CPU 111 and an input device 118 such as a keyboard and a mouse. The display controller 115 is connected to the display device 119 and controls display on the display device 119.

  The data reader / writer 116 mediates data transmission between the CPU 111 and the recording medium 120, and reads a program from the recording medium 120 and writes a processing result in the computer 110 to the recording medium 120. The communication interface 117 mediates data transmission between the CPU 111 and another computer.

  Specific examples of the recording medium 120 include general-purpose semiconductor storage devices such as CF (Compact Flash (registered trademark)) and SD (Secure Digital), magnetic recording media such as a flexible disk, or CD- An optical recording medium such as ROM (Compact Disk Read Only Memory) can be used.

  Note that the authentication device 10 according to the present embodiment can be realized not by using a computer in which a program is installed but also by using hardware corresponding to each unit. Further, the authentication apparatus 10 may be partially realized by a program and the remaining part may be realized by hardware.

  Part or all of the above-described embodiments can be expressed by (Appendix 1) to (Appendix 12) described below, but is not limited to the following description.

(Appendix 1)
An image acquisition unit that acquires a plurality of images that are continuous along the time series of the subject from the imaging device;
An authentication processing unit that determines whether or not each of the plurality of images matches a registered image registered in advance in the order that the plurality of images are captured by the imaging device, and performs authentication based on the determination result;
An image registration unit that registers, in association with the registered image, an image that has not been determined to match when the number of times that it has been determined to match is equal to or greater than a set number;
An authentication device comprising:

(Appendix 2)
In the case where the image registration unit calculates a match rate for any image determined to match in an image that has not been determined to match, and the calculated match rate is equal to or greater than a threshold value, An image that is not determined to match is registered in association with the registered image.
The authentication device according to attachment 1.

(Appendix 3)
The image registration unit determines whether the images that are not determined to match satisfy the setting requirements for image quality, and if the images are determined to match the setting requirements, An image that has not been determined is registered in association with the registered image.
The authentication device according to appendix 1 or 2.

(Appendix 4)
The subject is a person, and the registered image is a person's face image;
The authentication device according to any one of appendices 1 to 3.

(Appendix 5)
(A) acquiring a plurality of continuous images along the time series of the subject from the imaging device;
(B) determining whether each of the plurality of images matches a registered image registered in advance in the order taken by the imaging device, and performing authentication based on the determination result;
(C) when the number of times of continuous determination as matching is equal to or greater than a set number, registering an image that has not been determined to match in association with the registered image;
An authentication method comprising:

(Appendix 6)
In the case of the step (c), in the image that has not been determined to match, the matching rate for any image determined to match is calculated, and the calculated matching rate is equal to or greater than a threshold value To register an image that has not been determined to match with the registered image,
The authentication method according to attachment 5.

(Appendix 7)
In the step (c), it is determined whether the image that has not been determined to match the image quality satisfies the setting requirement for image quality. Register an image that is not determined to be associated with the registered image,
The method according to appendix 5 or 6.

(Appendix 8)
The subject is a person, and the registered image is a person's face image;
The authentication method according to any one of appendices 5 to 7.

(Appendix 9)
On the computer,
(A) acquiring a plurality of continuous images along the time series of the subject from the imaging device;
(B) determining whether each of the plurality of images matches a registered image registered in advance in the order taken by the imaging device, and performing authentication based on the determination result;
(C) when the number of times of continuous determination as matching is equal to or greater than a set number, registering an image that has not been determined to match in association with the registered image;
A program that executes

(Appendix 10)
In the case of the step (c), in the image that has not been determined to match, the matching rate for any image determined to match is calculated, and the calculated matching rate is equal to or greater than a threshold value To register an image that has not been determined to match with the registered image,
The program according to appendix 9.

(Appendix 11)
In the step (c), it is determined whether the image that has not been determined to match the image quality satisfies the setting requirement for image quality. Register an image that is not determined to be associated with the registered image,
The program according to appendix 9 or 10.

(Appendix 12)
The subject is a person, and the registered image is a person's face image;
The program according to any one of appendices 9 to 11.

  As described above, according to the present invention, it is possible to increase the number of master variations and improve the authentication accuracy without requiring manual processing. The present invention is useful for an entrance / exit management system and a monitoring system using face authentication. The present invention is also useful for a system for analyzing a flow line of a person for marketing.

DESCRIPTION OF SYMBOLS 10 Authentication apparatus 11 Image acquisition part 12 Authentication processing part 13 Image registration part 14 Storage part 15 Master database 110 Computer 111 CPU
112 Main Memory 113 Storage Device 114 Input Interface 115 Display Controller 116 Data Reader / Writer 117 Communication Interface 118 Input Device 119 Display Device 120 Recording Medium 121 Bus

Claims (5)

An image acquisition unit that acquires a plurality of images obtained by imaging a subject in time series from the imaging device ;
The features of each of the plurality of images are collated with the features of the registered images registered in advance in the order taken by the imaging device . If the features match, it is authenticated that the subject is associated with the registered image. An authentication processing unit;
In the authentication process, if the number of consecutive feature matches is greater than or equal to the set number, the match rate is calculated using the images that do not match the features and the images that match the features. And an image registration unit for registering an image whose features do not match in association with the registered image when the calculated matching rate is equal to or greater than a threshold value ;
An authentication device comprising: The image registration unit, an image feature do not match, if it meets the set requirements for quality, the image characteristic does not match, it is registered in association with said registered image,
The authentication device according to claim 1 . The subject is a person, and the registered image is a person's face image;
The authentication device according to claim 1 or 2 . (A) acquiring a plurality of images obtained by imaging the subject in time series from the imaging device ;
(B) The feature of each of the plurality of images is collated with the feature of a registered image registered in advance in the order taken by the imaging device, and if the features match, the subject associated with the registered image Authenticate , step,
(C) In the authentication process, when the number of times that features are continuously matched is equal to or greater than the set number, the images that are not matched and the images that match the features are matched using the captured images. Calculating a rate, and if the calculated match rate is equal to or greater than a threshold value , registering an image whose features do not match in association with the registered image;
An authentication method comprising: On the computer,
(A) acquiring a plurality of images obtained by imaging the subject in time series from the imaging device ;
(B) The feature of each of the plurality of images is collated with the feature of a registered image registered in advance in the order taken by the imaging device, and if the features match, the subject associated with the registered image Authenticate , step,
(C) In the authentication process, when the number of times that features are continuously matched is equal to or greater than the set number, the images that are not matched and the images that match the features are matched using the captured images. Calculating a rate, and if the calculated match rate is equal to or greater than a threshold value , registering an image whose features do not match in association with the registered image;
A program that executes

Images