JP2008107942A - Impersonation detection device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technology for detecting impersonation with a photograph or the like by simple processing without causing the enlargement of a device and degradation of convenience. <P>SOLUTION: At the substantially same timing as an in-camera 12 photographing a first image used for user authentication, an out-camera 13 photographs a second image. An impersonation detection device determines whether impersonation is performed when photographing the first image by analyzing the second image. The detector determines that impersonation is performed when a face is included in the second image, for instance. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a technique for detecting impersonation in a portable device having a user authentication function using a face image.

  In recent years, mobile devices (cell phones, PDAs (Personal Digital Assistants), mobile computers, etc.) have become more sophisticated and multifunctional, and data stored in the devices (personal information, image data, address book) Etc.) is also increasing in importance. Accordingly, security measures for preventing unauthorized use of portable devices have become important.

  As a security measure, facial recognition technology using a camera that is becoming standard equipment in portable devices has attracted attention. Personal authentication (face authentication) using face recognition technology is effective as a highly convenient security because it can capture a face from a natural motion and check the owner.

  On the other hand, there is a countermeasure against “spoofing” using a photograph or the like as a face recognition technique. Spoofing is an act in which another person tries to authenticate using his / her face photo or the like. In particular, recent mobile devices are equipped with megapixel-class cameras, and are equipped with functions such as autofocus and close-up functions, so it is possible to impersonate even small-sized photos such as snapshots and licenses. There is a risk of being able to do it. For example, if a license card or photo credit card is stolen with a mobile phone, the security lock of the mobile phone is released using those photos, and important data is read or electronic money is illegal May be used.

  As an idea for detecting impersonation, a method is known in which a user is requested to perform a predetermined action and whether the user has performed a movement as requested (see Patent Document 1). However, this method imposes an operation burden on the execution of the predetermined operation on the user, so that convenience that is an advantage of the face recognition technology is lost.

  As another idea, there are a method of taking multiple shots while changing the illumination conditions and confirming how the face is shaded (see Patent Document 2), and a method of examining whether it is a solid by shooting from various angles. is there. However, these methods require a plurality of illumination devices and a plurality of cameras or a camera moving mechanism, which makes the apparatus configuration large. In addition, since it is necessary to shoot many times by changing the illumination conditions and the shooting angle, it leads to a loss of convenience.

  As described above, since the conventional impersonation detection method leads to an increase in the size of the apparatus and a decrease in convenience, it has been difficult to mount the method on a portable device such as a mobile phone, a PDA, or a small computer.

Patent Documents 3 and 4 are techniques for using a camera of a portable device.
JP 2004-259255 A Japanese Patent Laid-Open No. 11-339048 JP 2003-317100 A JP 2004-147046 A

  The present invention has been made in view of the above circumstances, and the object of the present invention is to provide a technique capable of detecting impersonation by a photograph or the like with simple processing without causing an increase in the size of the apparatus or a decrease in convenience. It is to provide.

  In order to achieve the above object, the present invention adopts the following configuration.

  The present invention is mounted on a portable device having a user authentication function based on a face image, and detects impersonation by a photograph or the like. Here, “photograph and the like” includes not only hard prints but also images displayed on a display.

  An impersonation detection apparatus according to the present invention is a hardware resource of a portable device, and is a first imaging means provided to photograph one side of the portable device, and photographs the opposite side of the portable device. And a second imaging means provided to do so. The spoofing detection device analyzes the second image captured by the second imaging unit at substantially the same timing as when the first image used for user authentication is captured by the first imaging unit. Spoofing judging means for judging whether or not spoofing has been performed at the time of photographing the first image is provided.

  A person who tries to impersonate (hereinafter, referred to as “illegal user”) pays close attention to taking an image for user authentication (first image) by the first imaging means, but at this time, it is the opposite of the portable device. It seems that he / she does not care at all what is reflected in the second imaging means facing to the side. The present invention focuses on such psychological blind spots.

  Various specific determination processes of the impersonation determination means can be considered.

  For example, the impersonation determination unit may determine that impersonation has been performed when the first image is captured when a face is included in the second image.

  When an unauthorized user takes a picture of a mobile device (first imaging means) over a photo or the like facing him, the second imaging means faces the unauthorized user and illegally uses the second image. This is because the person's face is likely to be reflected.

  When a legitimate user takes a selfie for user authentication, the possibility that a third person's face appears in the second image is very low. However, it is also assumed that a face such as a passerby is included in the second image by chance.

  Therefore, the impersonation determination unit may determine that impersonation has been performed when the first image is captured when the second image includes a face and the size of the face is larger than a threshold value. Good.

  This is because even if a third party appears in the second image, the size of the face should be very small. By adding the face size to the determination condition as in this determination process, erroneous determination can be reduced as much as possible.

  Further, the impersonation determination unit may determine that impersonation has been performed when the first image is captured when the second image includes a face and the orientation of the face is the front.

This is because even if a third party appears in the second image by accident, the face orientation is considered to be mostly horizontal or diagonal. By adding the face orientation to the determination condition as in this determination process, erroneous determination can be reduced as much as possible. It is also preferable to add both the face size and the face orientation to the determination conditions.

  In addition, the impersonation determining means includes a third face in which a face is included in the second image, and a face of the same person as the face is taken by the first image pickup means after taking the first image. If it is also included in the image, it may be determined that impersonation has been performed when the first image is captured.

  In order to check whether the impersonation has succeeded in impersonation or to try to operate the mobile device, the unauthorized user reverses the orientation of the mobile device after taking a photograph or the like, and directs the first image pickup means toward the user. Probability is high. At this time, a phenomenon occurs in which the face that was captured in the second image at the time of user authentication appears in the third image captured by the first imaging means. The determination processing focuses on this phenomenon, and the impersonation determination reliability is improved by using not only the second image but also the third image for the determination.

  The determination processing described so far assumes a mode in which an unauthorized user performs shooting while holding a mobile device over a photograph or the like directed toward him. However, as an aspect of impersonation, an aspect in which an unauthorized user holds a portable device with the first imaging means facing him and holds a photograph or the like in front of the first imaging means to take a picture is also assumed. . In the case of this aspect, since the unauthorized user's face is not reflected on the second imaging means, the following determination process is effective.

  That is, the impersonation determination means includes a face that was not included in the first image included in the third image captured by the first imaging means after capturing the first image, and When the fourth image taken by the second imaging means coincides with the second image at substantially the same timing as when the third image is taken, spoofing is performed when the first image is taken. It is good to judge.

  Unauthorized users can take a picture in front of the first imaging means and take a picture, then immediately retreat the picture and check if the impersonation was successful or try to operate the mobile device High nature. At this time, although the face included in the image captured by the first imaging unit is a different person, there is almost no change in the image captured by the second imaging unit because the mobile device hardly moves. The phenomenon of not happening. The determination process focuses on this phenomenon.

  The present invention may be regarded as an impersonation detection device having at least a part of the above means, or may be regarded as a face authentication device or a portable device provided with an impersonation detection device. The present invention can also be understood as an impersonation detection method and a face authentication method including at least a part of the above processing, or a program for realizing such a method and a recording medium on which the program is recorded. Each of the above means and processes can be combined with each other as much as possible to constitute the present invention.

  For example, an impersonation detection method according to the present invention is an impersonation detection method in a portable device having a user authentication function based on a face image. The impersonation detection method is provided so that the portable device photographs one side of the portable device. 1 imaging means and 2nd imaging means provided so that the other side of the said portable device may be image | photographed, The said portable device image | photographed the 1st image used for user authentication by the said 1st imaging means. Processing, capturing the second image with the second imaging means at substantially the same timing as capturing the first image, and analyzing the second image, thereby capturing the first image. And a process of determining whether or not impersonation has been performed, and an impersonation detection method for executing.

An impersonation detection program according to the present invention is an impersonation detection program that is executed by a portable device having a user authentication function using a face image, and is provided so that the portable device captures one side of the portable device. First imaging means and second imaging means provided so as to photograph the opposite side of the portable device, and the first imaging device uses the first imaging means for user authentication. A process of capturing an image, a process of capturing a second image by the second imaging means at substantially the same timing as capturing the first image, and analyzing the second image to thereby analyze the first image. And a process for determining whether or not impersonation has been performed during the shooting of the image.

  According to the present invention, it is possible to detect impersonation by a photograph or the like with a simple process without causing an increase in the size of the apparatus or a decrease in convenience.

  Exemplary embodiments of the present invention will be described in detail below with reference to the drawings. In the following embodiments, a mobile phone is taken as an example, but the present invention can also be applied to other types of mobile devices (for example, PDAs, personal computers, game machines, etc.).

<Device configuration>
1A and 1B are views showing the appearance of a mobile phone, in which FIG. 1A shows the inside (operation side) of the mobile phone, and FIG. 1B shows the outside of the mobile phone.

  Inside the mobile phone 1, there are an operation unit 10 composed of numeric buttons, function buttons, etc., a display unit 11 composed of a liquid crystal display, etc., and an in-camera (inside camera) 12 as a first imaging means. Is provided. Further, an outside camera (outside camera) 13 as a second imaging unit is provided outside the mobile phone 1.

  The out camera 13 is provided in order to make it possible to take a picture with a mobile phone. On the other hand, the in-camera 12 is originally provided for photographing a speaker in a videophone. As described above, the mobile phone 1 is provided with the two cameras 12 and 13 having different uses, and the cameras 12 and 13 are arranged so as to photograph the opposite sides according to the use. Since the in-camera 12 is suitable for self-portrait, it is also used for shooting face images for user authentication.

  FIG. 2 is a block diagram showing a functional configuration of the mobile phone (spoofing detection device). The mobile phone 1 includes a first image acquisition unit 20, a second image acquisition unit 21, a face detection unit 22, a face collation unit 23, an impersonation determination unit 24, and the like as functions related to user authentication and impersonation detection. These functional elements are realized by the arithmetic processing unit executing software (program) and controlling hardware resources such as the cameras 12 and 13, the display unit 11, and the memory as necessary. However, part or all of these functional elements may be realized by a dedicated chip.

  The first image acquisition unit 20 has a function of controlling the in-camera 12 and capturing an image captured by the in-camera 12, and the second image acquisition unit 21 controls the out-camera 13 and is captured by the out-camera 13. It is a function to capture the image. The captured image is temporarily stored in the memory and used for user authentication and spoofing detection processing described later.

The face detection unit 22 has a function of detecting a human face from an image and specifying the position, size (size), orientation, and the like of the face. Any existing technique may be applied to the face detection processing by the face detection unit 22. For example, (1) a method for detecting a face by template matching using a reference template corresponding to the contour of the entire face, (2) a face by template matching based on a facial organ (eg, eyes, nose, ears). (3) A technique for detecting a vertex such as a head by chroma key processing and detecting a face based on the vertex, and (4) detecting an area close to the skin color and detecting the area as a face. And (5) learning using a teacher signal using a neural network and detecting a face-like area as a face. Instead of automatically detecting a face, a person may specify the position and size of the face. When a plurality of human faces are detected from the image, the face to be processed may be determined based on predetermined criteria such as the size, orientation, and position of the face. Therefore, the face matching process may be interrupted.

  The face matching unit 23 is a function that compares two faces and determines whether or not they are the same person's face. The face collation unit 23 can collate the face detected by the face detection unit 22 with the face of a registered user registered in the memory in advance, or can collate two faces detected from different images. it can. Any existing technique may be applied to the face matching process by the face matching unit 23. For example, the similarity is calculated by comparing one or a plurality of feature amounts respectively extracted from two faces to be collated, and the two faces are determined by whether or not the similarity satisfies a predetermined condition. There is a method for determining whether or not they belong to the same person.

  The impersonation determination unit 24 is a function that determines whether impersonation has been performed when an image used for user authentication is captured by the in-camera 12. Specific processing will be described later.

<User authentication>
The mobile phone 1 has a user authentication function using a face image. When the user tries to use the function with the security lock, the user authentication function is automatically activated to request the user to capture a face image.

  FIG. 3 shows a correct photographing posture for user authentication. That is, the correct photographing posture is to hold the mobile phone 1 so that the in-camera 12 faces his / her face. Images are captured from the in-camera 12 at regular time intervals, and the images are sequentially displayed on the display unit 11. Therefore, the user 30 can adjust the position, size, orientation, etc. of the face while viewing the displayed images. it can. When the user 30 presses the shutter button of the operation unit 10, an image for user authentication is taken by the in-camera 12.

  FIG. 4 is a flowchart illustrating an example of a user authentication routine. First, the face detection unit 22 detects a face from an image for user authentication taken by the in-camera 12 (step S500). If a face cannot be detected (step S501; NO), a rejection determination is made assuming that the authentication has failed (step S505). On the other hand, when a face can be detected (step S501; YES), the face collation unit 23 compares the detected face with the registered user's face (step S502). When it is determined that the two faces belong to different people (step S503; NO), a rejection determination is made on the assumption that the authentication has failed (step S505). If it is determined that the two faces belong to the same person (step S503; YES), a permission determination is made on the assumption that the authentication has been successful (step S504).

  There is also a simple user authentication method in which the face matching process by the face matching unit 23 is omitted. FIG. 5 is a flowchart showing a modification of the user authentication routine. In the processing example of FIG. 5, if a face can be detected from an image for user authentication photographed by the in-camera 12 (steps S400, S401; YES), permission determination is made (step S402), and if a face cannot be detected (step S402). S401; NO), rejection determination (step S403). This user authentication routine is low in reliability because it does not perform face matching, but has the advantage of high speed processing. Therefore, the highly reliable routine of FIG. 4 and the high-speed routine of FIG. 5 may be selectively used according to the required security level.

  If the security lock is released only when the result of the user authentication is a permission determination, unauthorized use of the mobile phone 1 by a person other than the authorized user can be prevented.

  However, user authentication using such a face image is not perfect. This is because there is a problem of so-called “spoofing” in which authentication is performed using a face photograph of a legitimate user (person). Therefore, the cellular phone 1 according to the present embodiment executes impersonation detection processing as described below in conjunction with user authentication, and prohibits the release of the security lock when it is determined that impersonation has been performed during user authentication.

<Impersonation detection>
Hereinafter, processing examples 1 to 4 will be exemplified as specific examples of the spoofing detection process.

(Processing example 1)
FIG. 6 shows one mode of impersonation. As shown in FIG. 6, when the unauthorized user 40 takes a picture 41 facing the user and holds the in-camera 12 of the mobile phone 1, the out-camera 13 faces the unauthorized user 40. Therefore, if photographing with the in-camera 12 and photographing with the out-camera 13 are performed simultaneously, as shown in FIG. 7A, there is a high possibility that the face of the unauthorized user 40 can be photographed. On the other hand, when a legitimate user 30 takes a selfie as shown in FIG. 3, the possibility that a third party's face appears in the out-camera image is very low (see FIG. 7B).

  Processing example 1 focuses on this phenomenon. Now, the flow of the processing example 1 will be described along the flowchart of FIG.

  After the user authentication function is activated, when the shutter button of the operation unit 10 is pressed, an image for user authentication is taken by the in-camera 12 (step S100). At this time, shooting is also performed with the out-camera 13 at substantially the same timing as shooting with the in-camera 12 (step S101). Note that the steps S100 and S101 may be interchanged as long as shooting with both cameras is performed at substantially the same timing.

  Next, a spoofing determination routine is executed using the out-camera image (second image) photographed in step S101 (step S102).

  FIG. 9 shows a flowchart of an impersonation determination routine according to Process Example 1. First, the face detection unit 22 tries to detect a face from the out-camera image (step S200). Then, when a face is included in the out-camera image (step S201; YES), the impersonation determination unit 24 determines that impersonation has been performed at the time of shooting in step S100 (step S202), and no face has been detected. In this case (step S201; NO), it is determined that impersonation has not been performed (step S203).

  In the case of a determination result that impersonation has been performed (step S103 in FIG. 8; YES), the processing is immediately terminated. On the other hand, in the case of the determination result that it is not impersonation (step S103; NO), the user authentication routine is executed using the in-camera image (first image) photographed in step S100 (step S104). The specific process of the user authentication routine is the same as that shown in FIG. When the permission determination is made in step S104 (step S105; YES), the security lock is released (step S106), and the mobile phone 1 can be operated.

According to the above processing example 1, since the impersonation of the mode shown in FIG. 6 can be detected, unauthorized use of the mobile phone 1 by a person other than a legitimate user can be prevented. In addition, since a special illumination device or imaging device is not required for taking an image, there is no fear of increasing the size of the device or reducing convenience when implementing impersonation detection.

(Processing example 2)
When a legitimate user takes a selfie, the possibility that a third party's face appears in the out-camera image is very low. However, when user authentication is performed in a crowded place or on a train, it is assumed that the faces of surrounding people are accidentally reflected in the out-camera image. In this case, in Processing Example 1, although a legitimate user performs user authentication, it is erroneously determined as “spoofing”.

  Therefore, in the processing example 2, not only the presence / absence of the face but also the size of the face is added to the determination condition. FIG. 10 shows a flowchart of an impersonation determination routine according to Processing Example 2. Since other processes are the same as those in Process Example 1, the description thereof is omitted.

  As shown in FIG. 10, first, the face detection unit 22 tries to detect a face from the out-camera image (step S300). If a face is included in the out-camera image (step S301; YES), the impersonation determination unit 24 checks whether the detected face size is larger than a predetermined threshold value (step S301). S302). If the face size is larger than the threshold (step S302; YES), it is determined to be impersonation (step S303). If the face cannot be detected (step S301; NO), the face size is equal to or smaller than the threshold. If it is (step S302; NO), it is determined that it is not impersonation (step S304).

  Even if a third party accidentally appears in the out-camera image, the face size should be very small. Therefore, by adding the face size to the determination condition as in Processing Example 2, it is possible to reduce erroneous determination as much as possible.

  Note that the orientation of the face may be taken into account as a determination condition for distinguishing between an unauthorized user and a third party who accidentally appears. In other words, if the face is facing the front, it is determined to be impersonated by an unauthorized user, and if the face is facing diagonally or sideways, it is determined that it is not impersonating. This method can also reduce erroneous determinations.

(Processing example 3)
An unauthorized user is likely to turn the direction of the mobile phone 1 and point the in-camera 12 toward the user after taking a picture as shown in FIG. This is for confirming whether photography or spoofing has been successful, or for attempting to operate the mobile phone 1. When an unauthorized user takes such an action, as shown in FIG. 11, the face of the unauthorized user 40 that appears in the out-camera image at the time of user authentication appears in the in-camera image at a certain timing. A phenomenon occurs. Such a phenomenon cannot occur in the case of authentication by a legitimate user.

  Processing example 3 focuses on this phenomenon. Then, the flow of the process example 3 is demonstrated along the flowchart of FIG.

  When the shutter button of the operation unit 10 is pressed after the user authentication function is activated, an image for user authentication is captured by the in-camera 12 and also by the out-camera 13 (steps S600 and S601). These processes are the same as those in FIG.

Next, a user authentication routine is executed using the in-camera image (first image) taken in step S600 (step S602). The specific process of the user authentication routine is the same as that shown in FIG. If a rejection determination is made in step S602 (step S603; NO), the process ends. On the other hand, when the permission determination is made in step S602 (step S603; YES), the security is released (step S604), and the process proceeds to step S605.

  After waiting for a predetermined time in step S605, the process of step S606 is executed. This wait is provided to wait for the completion of the reversing operation of the mobile phone 1. Instead of the wait process, when there is some user operation on the operation unit 10, the process may proceed to step S606. Alternatively, the mobile phone 1 may be provided with detection means such as a posture sensor, a tilt sensor, or an acceleration sensor, and the reverse of the mobile phone 1 may be detected by the detection means.

  In step S606, a spoofing determination routine is executed using the out-camera image (second image) taken in step S601.

  FIG. 13 shows a flowchart of an impersonation determination routine according to Processing Example 3. First, the face detection unit 22 tries to detect a face from the out-camera image (step S700). When a face is detected, a new image is taken by the in-camera 12 (step S702), and the face is detected from the newly taken in-camera image (third image) (step S703). . The processes in steps S702 and S703 are repeated until a face is detected from the in-camera image (step S704).

  Subsequently, the face matching unit 23 compares the face detected from the out-camera image (second image) in step S700 with the face detected from the in-camera image (third image) in step S703 (step S705). ). When it is determined that the two faces belong to the same person (step S706; YES), the impersonation determination unit 24 determines that impersonation has been performed during user authentication (step S707).

  When a face cannot be detected from the out-camera image (step S701; NO) and when the two faces are not from the same person (step S706; NO), the impersonation determination unit 24 It is determined that impersonation has not been performed (step S708).

  In the case of the determination result that the impersonation has been performed (step S607 in FIG. 12; YES), the security lock is activated again, and the operation of the mobile phone 1 becomes impossible (step S608).

  As a result, unauthorized use due to impersonation can be prevented. Moreover, the processing example 3 improves the reliability of impersonation determination by using not only the out-camera image but also the in-camera image for the determination.

(Processing example 4)
As shown in FIG. 6, the processing examples 1 to 3 assume a mode in which an unauthorized user 40 takes a picture while holding the mobile phone 1 over a photo 41 directed toward the user. However, as a form of impersonation, as shown in FIG. 14, an unauthorized user 40 holds the mobile phone 1 with the in-camera 12 facing him and holds a photo 41 in front of the in-camera 12 to take a picture. A mode of performing is also envisaged. In the case of this aspect, since the face of the unauthorized user 40 is not reflected on the out camera 13, the processing examples 1 to 3 cannot be used.

  Therefore, in the processing example 4, processing that focuses on the following phenomenon is performed.

An unauthorized user may take a picture in the manner shown in FIG. 14 and immediately withdraw the picture, confirm whether the picture has been successfully taken or impersonated, or may attempt to operate the mobile phone 1. high. Then, as shown in FIG. 15, before and after the photo is retreated, the face included in the in-camera image is a different person, but the mobile phone 1 hardly moves, so the out-camera image is almost non-existent. The phenomenon of no change occurs. In the case of authentication by a legitimate user, the possibility that the above phenomenon occurs is extremely low. This is because even if a legitimate user authenticates himself / herself and makes another person operate the mobile phone 1, the mobile phone 1 is almost always moved.

  FIG. 16 shows a flowchart of an impersonation determination routine according to Processing Example 4. This routine is executed in step S606 of the flowchart of FIG.

  First, a new image is captured by the in-camera 12 (step S800), and a face is detected from the newly captured in-camera image (third image) (step S801). Steps S800 and S801 are repeated until a face is detected from the in-camera image (step S802).

  Next, the face detected from the user authentication image (first image) in step S400 (or S500) is compared with the face detected in step S801 (step S803). If the two faces belong to the same person (step S804; YES), it is determined that the face is not impersonation (step S809). If the two faces are not of the same person, that is, if the face included in the in-camera image has been changed to that of another person (step S804; NO), the process proceeds to step S805.

  In step S805, a new image is taken by the out camera 13. Then, the newly photographed out-camera image (fourth image) is compared with the out-camera image (second image) at the time of user authentication photographed in step S601 (step S806). Here, if it is determined that the two images match (step S807; YES), it is determined to be impersonation (step S808). Note that in the comparison processing in step S806, it is not required that the two images are completely identical, and if the two images can be evaluated as having been taken from substantially the same position, it is determined as “match”. . Any existing technique may be applied to the image comparison process.

  According to this processing example 4, impersonation of the aspect shown in FIG. 14 can be detected, and unauthorized use of the mobile phone 1 can be suppressed.

  By the way, the above-mentioned impersonation detection may cause an erroneous determination when an unauthorized user takes an unexpected behavior. For example, processing examples 1 to 3 cannot detect impersonation when an unauthorized user is outside the angle of view of the out-camera 13, and processing example 4 greatly moves the mobile phone 1 when the unauthorized user rejects the photo. Impersonation cannot be detected in the event that However, the impersonation detection described above is extremely advantageous in terms of mounting because no additional hardware is required and processing is simple. Therefore, it can be preferably employed when good operability and low cost are more important than high security. Note that false detection can be reduced in combination with other processing.

  The above embodiment is merely an example of the present invention. The scope of the present invention is not limited to the above embodiment, and various modifications can be made within the scope of the technical idea.

1A and 1B are views showing the appearance of a mobile phone, FIG. 1A is a view showing the inside (operation side) of the mobile phone, and FIG. 1B is a view showing the outside of the mobile phone. FIG. 2 is a block diagram showing a functional configuration of the mobile phone. FIG. 3 is a diagram illustrating a correct photographing posture for user authentication. FIG. 4 is a flowchart illustrating an example of a user authentication routine. FIG. 5 is a flowchart showing a modification of the user authentication routine. FIG. 6 is a diagram illustrating one mode of impersonation. FIG. 7A is a diagram illustrating an example of a captured image in the case of impersonation, and FIG. 7B is a diagram illustrating an example of a captured image in the case of a legitimate user. FIG. 8 is a flowchart illustrating processing example 1 of impersonation detection. FIG. 9 is a flowchart of an impersonation determination routine according to Processing Example 1. FIG. 10 is a flowchart of an impersonation determination routine according to Processing Example 2. FIG. 11 is a diagram illustrating an example of a captured image in the case of impersonation. FIG. 12 is a flowchart illustrating a third example of impersonation detection processing. FIG. 13 is a flowchart of an impersonation determination routine according to Processing Example 3. FIG. 14 is a diagram showing another mode of impersonation. FIG. 15 is a diagram illustrating an example of a captured image in the case of impersonation in FIG. FIG. 16 is a flowchart of an impersonation determination routine according to Processing Example 4.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Mobile phone 10 Operation part 11 Display part 12 In camera 13 Out camera 20 1st image acquisition part 21 2nd image acquisition part 22 Face detection part 23 Face collation part 24 Impersonation determination part 30 User 40 Unauthorized user 41 Photograph

Claims (8)

An impersonation detection device in a portable device having a user authentication function using a face image,
First imaging means provided so as to photograph one side of the portable device;
A second imaging means provided to photograph the opposite side of the portable device;
When the first image is captured, the second image captured by the second image capturing unit is analyzed at substantially the same timing as when the first image used for user authentication is captured by the first image capturing unit. Impersonation determination means for determining whether or not impersonation has been performed;
Impersonation detection device comprising   The impersonation detection device according to claim 1, wherein the impersonation determination unit determines that impersonation has been performed when the first image is captured when a face is included in the second image.   The impersonation determination means determines that impersonation has been performed when the first image is captured when the second image includes a face and the size of the face is larger than a threshold value. Impersonation detection device described in 1.   The said impersonation determination means determines that the impersonation was performed at the time of imaging | photography of the said 1st image, when the face is contained in the said 2nd image and the direction of the face is a front. Impersonation detection device.   The impersonation determination unit includes a third image in which a face is included in the second image, and a face of the same person as the face is captured by the first imaging unit after the first image is captured. The impersonation detection device according to claim 1, wherein when the first image is captured, it is determined that impersonation has been performed.   The impersonation determination means includes a face that was not included in the first image included in a third image captured by the first imaging means after capturing of the first image, and the first image When the fourth image captured by the second imaging unit coincides with the second image at substantially the same timing as when three images are captured, spoofing was performed when the first image was captured. The impersonation detection apparatus of Claim 1 which determines. An impersonation detection method in a mobile device having a user authentication function based on a face image,
The portable device includes first imaging means provided so as to photograph one side of the portable device, and second imaging means provided so as to photograph the opposite side of the portable device. ,
The portable device is
Processing for capturing a first image used for user authentication by the first imaging means;
A process of taking a second image by the second imaging means at substantially the same timing as taking the first image;
A process of determining whether or not impersonation has been performed at the time of photographing the first image by analyzing the second image;
Impersonation detection method to execute. An impersonation detection program executed on a portable device having a user authentication function based on a face image,
The portable device includes first imaging means provided so as to photograph one side of the portable device, and second imaging means provided so as to photograph the opposite side of the portable device. ,
In the portable device,
Processing for capturing a first image used for user authentication by the first imaging means;
A process of taking a second image by the second imaging means at substantially the same timing as taking the first image;
A process of determining whether or not impersonation has been performed at the time of photographing the first image by analyzing the second image;
Impersonation detection program that executes

Images