Device and procedure for generating a biometric image of a person's face

Abstract

The invention relates to a device (100) for generating a biometric image of a person's face using a semi-transparent mirror (103), the surface of which reflects on the front face (104) and transmits on the rear face (109), which is opposite the front face (104), having at least one display device (110) arranged on the rear face (109) of the semi-transparent mirror (103) and directed towards the transmitting surface of the rear face (109) of the semi-transparent mirror (103), and having at least one camera (101) for capturing an image of the person's face, the camera beam of which extends through the semi-transparent mirror (103) to the person's face, and having a control unit (102) and an output unit (111) designed to emit information or an instruction as a visual or audible guide for the user to generate an augmented reality.The invention also relates to a method for generating a biometric image.

Description

[0001] The invention relates to a device and a method for generating a biometric image of a person's face.

[0002] Photographs for identification or security documents, such as passports or other travel documents, must meet biometric requirements. These photographs are often taken by a photographer. The photographer provides instructions to ensure the image meets biometric requirements, such as the correct orientation of the person to the camera, a closed mouth, and open eyes. However, the application for such a security or identification document typically takes place at a government office, a separate location from the photographer. Therefore, applying for a security or identification document requires the individual to travel to two different locations: one for the photograph and the other for the application. Furthermore, current passport photo booths require considerable space and are thus unsuitable for use in confined areas.

[0003] Furthermore, as part of automated access control, biometric images of individuals are taken and compared with a biometric image from an identification document. To enable the taking of a biometric image, the individual is given instructions or informed of various requirements. These instructions / information can sometimes be ambiguous, especially if they are not understood due to a language barrier, making the process occasionally time-consuming.

[0004] In US 2020 / 0 105 158 A1, a digitized "makeup mirror" is described that uses augmented reality to give instructions for facial treatment.

[0005] In DANIEL BESSERER ET AL: "Fitmirror" MULTIMODAL ANALYSES ENABLING ARTIFICIAL AGENTS IN HUMAN-MACHINE INTERACTION, ACM, 2 PENN PLAZA, SUITE 701 NEW YORK NY 10121-0701 USA, November 12, 2016 (2016-11-12), pages 48-55, XP058308178, DOI: 10.1145 / 3011263.3011265; ISBN: 978-1-4503-4562-2, a concept for a "smart mirror" is described, supplemented by a "balance board" and a "Kinect camera." It is designed to recognize a user standing on the "balance board" and guide them through fitness exercises. CN111434280A also describes such a "smart mirror".

[0006] WO 2015 / 058948 A1 describes an access control station where security checks can be carried out. Data from the person requesting access is recorded, which may include biometric data. A recording device is concealed behind a mirror. The access control station also includes a document scanner and a printer for affixing stamps to presented identification documents. Information can be displayed on the semi-transparent mirror via a display device. This allows the person to be prompted to place a document or position their face as part of a user guidance system. Voice output is also possible. The concealed recording device includes a camera. A control unit is also present.

[0007] EP 3 432 065 A1 also describes a device for capturing facial images using a terminal. The terminal includes a reader for scanning a passport. Camera lenses are located behind a mirror, and these lenses capture the user's facial image for comparison with the scanned passport photo. The camera lenses can be positioned between display devices.

[0008] German patent DE102018132475A1 describes a device for capturing biometric features of a person's face. It includes an illuminated back panel to "block out" shadows in the images. A deflecting mirror is located in the light path between the camera and the face.

[0009] In DE 10 2014 109 682 A1 a mobile terminal for recording biometric data is described, which is designed in the form of a portable case.

[0010] US patent 2007 / 0047774A1 describes a device for capturing biometric features of a person's face, which is equipped with a display to give instructions to the user.

[0011] Therefore, the object of the present invention is to provide a compact device and a method for generating a biometric image of a person's face, which enables simplified user guidance.

[0012] This problem is solved by a device having the features of claim 1 and by a method having the features of claim 12. Advantageous embodiments with expedient further developments of the invention are specified in the dependent claims.

[0013] The device for generating a biometric image of a person's face, using a semi-transparent mirror whose front surface reflects incident light and whose rear surface opposite the front surface transmits incident light, includes at least one display device arranged on the rear of the semi-transparent mirror, directed at the transmitting surface of the rear of the semi-transparent mirror. Furthermore, at least one camera is provided for capturing the person's face, the optical path of which passes through the semi-transparent mirror to the person's face. In particular, a control unit and an output unit are provided and configured to output information or instructions as optical or acoustic user guidance for generating augmented reality.

[0014] Augmented reality refers to a computer-aided enhancement of reality perception. The information or input can therefore be any signals perceptible through human senses, such as visual or auditory signals. For example, in augmented reality, images or sequences of images, i.e., videos, are supplemented with computer-generated additional information or virtual objects through overlays or superimpositions. With a semi-transparent mirror—also called a Venetian mirror—incident light is reflected on the reflective surface; that is, an object positioned in front of the reflective surface of the semi-transparent mirror is mirrored.The back side of the transmitting mirror, however, transmits the incident light, allowing a person positioned on the transmitting side of the semi-transparent mirror to see through it. The advantage of this device lies in the fact that the optical or acoustic user guidance used to create an augmented reality experience makes it easier to inform the user of the biometric requirements and follow the instructions. This makes capturing or generating a biometric image more user-friendly. Furthermore, the device has a compact design. It can be mounted on a wall or stand, or placed on a pedestal, making it suitable for use in smaller rooms, particularly in government offices.This allows individuals to have their biometric image taken at the same location where they apply for their identification document. The device's compact design also makes it possible to install a large number of these devices in a small space for automated access control.

[0015] In a preferred embodiment, the display device is arranged on the rear side of the camera facing away from the semi-transparent mirror. This allows, as in an example not covered by the invention, the display device to have only one display. Furthermore, it enables the person to clearly see both their own reflection on the semi-transparent mirror and the information, instructions, or images displayed on the display device and projected onto the semi-transparent mirror.

[0016] In an alternative embodiment, it is preferred that the camera is arranged on the rear side of the display device facing away from the semi-transparent mirror, and that the beam path runs through a display-device-free area and through the semi-transparent mirror to the person's face. In this embodiment, the display device is therefore arranged between the camera and the semi-transparent mirror.

[0017] In another alternative embodiment, a deflecting mirror is provided by means of which an image displayed on the display device is deflected onto the partially transparent mirror. This makes it possible to position the display device at any point within the device.

[0018] In order to keep the device as compact as possible, the invention provides for a deflecting mirror to fold the beam path running between the camera and the person's face.

[0019] For simplified operation of the device, it is preferred that the semi-transparent mirror has a touch-sensitive touchscreen on its reflective front surface. Alternatively or additionally, a control element can also be attached to or next to the housing.

[0020] It is advantageous that the semi-transparent mirror is framed, and that lighting elements are attached to the frame, forming a light strip. This light strip serves to illuminate the face of the person well when creating the biometric image.

[0021] In this context, it is advantageous if the light strip also extends along the entire frame in a frame-like manner, as this provides particularly good illumination.

[0022] It is further preferred that the lighting elements can be controlled separately and are controlled in order to illuminate the lighting elements of the light bar with different intensities and / or sequentially. In this way, it is possible to verify, based on the presence or absence of a shadow on the person's face, whether the person in front of the camera is a real, living person or whether only a photograph of a person is being held in front of the camera (implementation of so-called PAD = "presentation attack detection").

[0023] Alternatively or additionally, it is advantageous if the light strip also runs in a frame-like fashion and is embedded flush within the frame. This design makes the device particularly compact while still allowing for good illumination of the face.

[0024] The device is easy to operate because the semi-transparent mirror itself is designed as a touch-sensitive touchscreen.

[0025] Alternatively, the display device (or all display devices) can be designed as a touch-sensitive touchscreen, and the semi-transparent mirror can be at least partially electrically conductive and in contact with the display device in such a way that the touchscreen can be operated through the semi-transparent mirror. In this way, the entire device can be designed to be very space-saving and compact.

[0026] Alternatively or additionally, the device can be controlled by gesture recognition or voice control to simplify user guidance.

[0027] In a preferred embodiment, the output unit is formed as the at least one display device, wherein the information or instruction is displayed as visual user guidance on the display device and transmitted through the semi-transparent mirror, thus being visible. The output unit is therefore preferably formed as the at least one display device. This enables the simplest possible visual / optical user guidance.

[0028] The at least one camera can be configured as a two-dimensional camera, although it is particularly preferred if the at least one camera is configured as a three-dimensional camera. This enables the recording of three-dimensional images and also allows the person to orient themselves in space relative to the camera.

[0029] Furthermore, the device can also have a plurality of cameras, in particular a plurality of two-dimensional cameras, but also a plurality of three-dimensional cameras that optically detect in the visible range and / or in the infrared range.

[0030] To effectively display the user interface for generating augmented reality, it is preferred to have multiple display devices arranged in such a way that a display-device-free area, through which the camera's beam path passes, is formed between them. This ensures that the face and its biometric features can be captured with sufficient accuracy. The display-device-free area can be a recess in the display device or arranged between multiple display devices. Furthermore, particularly to protect the optics, it is also possible for a glass pane to be associated with or integrated into the display-device-free area.

[0031] Furthermore, it is preferred if the semi-transparent mirror is at least partially framed, and if at least one sensor is attached to or within the frame. The at least one sensor can be a loudspeaker and / or a lighting element and / or another camera, in particular an overview camera and / or an infrared camera and / or a thermal imaging camera and / or a microphone for voice control.

[0032] In this context, it is advantageous if the output unit or an additional output unit is designed as a loudspeaker, enabling acoustic user guidance to create an augmented reality experience. This allows blind or visually impaired individuals, in particular, to follow instructions and information.

[0033] To enable the device to be used by people of different heights while maintaining a compact design, it is advantageous for the semi-transparent mirror, the at least one display device, and the at least one camera to be arranged in a single housing, and for the housing to be mounted in a height-adjustable and / or pivotable manner. In an alternative embodiment, the at least one display device and the at least one camera can be arranged in a single housing, and the housing can be mounted in a height-adjustable manner relative to the mirror. Alternatively, only the optical system can be mounted in a height-adjustable manner relative to the mirror. This would have the advantage that no moving parts are accessible to the user, thus reducing the risk of injury.

[0034] The method for generating a biometric image of a person's face using a device described above includes, in particular, the following steps: A) Positioning a person in front of the reflective front of the semi-transparent mirror in such a way that the person's face is reflected on the reflective surface, B) Capturing the person's face by at least one camera through the semi-transparent mirror and generating a first image or a first sequence of images, C) Check whether the image or sequence of images produced by means of at least one camera meets at least one predefined requirement, D) In the event of a deviation of the generated first image or image sequence from at least one predefined requirement, an information or instruction is output in the output unit as optical or acoustic user guidance to generate an augmented reality, and E) the taking of another image or sequence of images by the camera, especially after correction or elimination of the deviation in section D).

[0035] This enables user guidance that is easy for the individual to understand using augmented reality. Specifically, the process is repeated until all predefined requirements, particularly the requirements for a biometric image, are met. The process can be carried out either during the creation of a biometric image or as part of automated access control.

[0036] Furthermore, the method prefers to include enhanced liveness detection of the user. This enhanced liveness detection can be achieved, for example, using motion profiles or infrared measurements with an infrared camera.

[0037] In this context, it is also advantageous if the identification document is read by placing it on the reflective surface of the semi-transparent mirror or on a reading device. The control unit can then compare the generated biometric image with the biometric image of the identification document and grant access if the images / biometric characteristics match; if there is a discrepancy, access is denied. Alternatively, it is also advantageous if the identification document is presented within the depth of field of the person in relation to the camera, i.e., the person holds the identification document within the area captured by the optical camera. This allows the camera to optically capture the identification document, and the control unit can then compare the generated biometric image with the biometric image of the identification document.

[0038] To enable the device to capture images of people of varying heights, it is preferred that the device's housing be height-adjusted relative to the person until the person's face can be fully captured by the camera. In particular, it is preferred that the device's housing be height-adjusted relative to the person until the person's face is centered within the area optically detectable by the camera, i.e., centered on the optical axis. This reduces optical aberrations.

[0039] Furthermore, it is particularly advantageous if the control unit determines the position of multiple reference points in at least one image or image sequence relative to the camera. The biometric requirements of the image can be verified based on these reference points and their positions. Additionally, the user's size and biometric measurements can be performed using these reference points, especially in relation to the camera and the surrounding space.

[0040] In this context, it is further preferred if the control unit assigns the output of the information or instruction to at least one of the reference points, such that the information or instruction, i.e., for example, text, a symbol, or the face itself, is displayed on the display device in the area of ​​the point on the image corresponding to the reference point and / or on the reflection of the person's face produced by the semi-transparent mirror. This enables visual user guidance to create an augmented reality.

[0041] In particular, the predetermined reference points can be the mouth and / or the corners of the mouth and / or the eyes and / or the corners of the eyes and / or the chin and / or the nose and / or the tip of the nose. This makes it possible to assign requests / information to the reference points on the person's face, i.e., to display them where they are relevant. For example, an instruction to open the eyes can be assigned to the reference points of the eyes or the corners of the eyes and thus displayed in the area of ​​the eyes / corners of the eyes on the image or on the mirror image. The instruction to close the mouth can, in turn, be assigned to the reference point associated with the mouth or corners of the mouth and displayed there.

[0042] The features and combinations of features mentioned above in the description, as well as those subsequently mentioned in the figure description and / or shown in the figures alone, are not only usable in the combinations specified, but also in other combinations or on their own. Therefore, embodiments not explicitly shown and explained in the figures are also to be considered as encompassed and disclosed by the invention.

[0043] Further advantages, features and details of the invention will become apparent from the claims, the following description of preferred embodiments, and the drawings. These show: Fig. 1 a schematic side view of the device for generating a biometric image, Fig. 2 a rear view of the device for generating a biometric image Figure 1; and Fig. 3 a front view of the device for generating a biometric image from Figure 1 .

[0044] Figure 1Figure 1 shows a side view of the device 100 for generating a biometric image of a person's face. The device 100 has a partially transparent, in this case a semi-transparent, mirror 103, the surface of which reflects incident light on the front 104 and transmits incident light on the rear 109 opposite the front 104. Furthermore, at least one display device 110 is arranged on the rear 109 of the semi-transparent mirror 103 and is directed towards the transmitting surface of the rear 109 of the semi-transparent mirror 103.Furthermore, at least one camera 101 for recording the person's face is arranged in the housing 107. The camera 101 is located on the rear side of the display device 110, facing away from the semi-transparent mirror 103, and its beam path runs through a display-device-free area 112 and through the semi-transparent mirror 103 to the person's face. The camera 101 and the display device 110 are connected to a control unit 102 via a communication link 113. An output unit 111 is also provided, which is likewise in communication link with the control unit 102. The control unit 102 and the output unit 111 are configured to output information or instructions as optical or acoustic user guidance to generate augmented reality.Furthermore, the device 100 additionally features a deflecting mirror 105 for folding the beam path 106 running between the camera 101 and the person's face. This enables a compact design of the device 100. The semi-transparent mirror 103, the at least one display device 110, the at least one camera 101, the deflecting mirror 105, and the control unit 102 are arranged in a housing 107, which is height-adjustable and pivotable. The device 100 can alternatively or additionally be mounted on a wall (not shown).

[0045] The device 100 optionally includes a reader 108. An identification document can be placed on the reader and thus read. A touch-sensitive touchscreen is preferably assigned to the reflective front surface 104 of the semi-transparent mirror 103, with which the device 100 can be operated. Alternatively or additionally, the device 100 can also be controlled via gesture recognition and / or voice control. In this context, the output unit 111 is formed as the at least one display device 110. The information or instruction is thus displayed as visual user guidance on the display device 110 and transmitted through the semi-transparent mirror 103, making the information / instruction perceptible or visible to the person / user on the semi-transparent mirror 103. The at least one camera 101 is, in this context, a three-dimensional camera.

[0046] Figure 2Figure 1 shows a rear view of the device 100 for generating a biometric image. It is evident that the device 100 preferably has several display devices 110, in this case two display devices 110, which are arranged relative to each other such that the display-device-free area 112, through which the beam path 106 of the camera 101 runs, is formed between them, i.e., runs between them. Furthermore, it can be seen from the Figure 2The semi-transparent mirror 103 is framed by a frame 114. A plurality of sensors (not shown in detail) can be attached to the frame 114. These sensors can be configured as additional cameras, in particular as an overview camera in the visual wavelength range and / or an infrared camera for liveness detection and / or as an additional two- or three-dimensional camera. Furthermore, the sensors can also function as loudspeakers. For example, an additional output unit 111, configured as a loudspeaker, can be provided so that information or instructions for acoustic user guidance can be output to create augmented reality. Lighting elements, in particular light strips, can also be attached to the frame 114 to ensure optimal illumination of the person's face. Microphones can also be provided for speech recognition.

[0047] Figure 3 shows a front view of the device 100, in which the two display devices 110 are represented by dashed lines.

[0048] The method for generating a biometric image of a person's face using the device 100 described above comprises, in particular, the following steps: First, the person is positioned in front of the reflective front surface 104 of the semi-transparent mirror 103 such that the person's face is reflected on the reflective surface 104. The housing 107 of the device 100 may be adjusted in height until the face can be fully imaged by the camera and is thus completely reflected on the reflective front surface 104 of the semi-transparent mirror 103. Subsequently, the person's face is captured by the at least one camera 101 through the semi-transparent mirror 103, thereby generating a first image or a first sequence of images.The generated first image or image sequence is checked by the control unit 102 to ensure it meets the specified requirements, particularly the requirements for a biometric image. If the generated first image or image sequence deviates from at least one of the specified requirements, corresponding information or instructions are output to the output unit 111 as visual or acoustic user guidance for generating augmented reality. Finally, another image or image sequence is generated by the camera 101.

[0049] The control unit 102 preferably determines the position of a plurality of reference points in the at least one image or image sequence with respect to the camera 101. The reference points are preferably the mouth and / or the corners of the mouth and / or the eyes and / or the corners of the eyes and / or the chin and / or the nose and / or the tip of the nose. Based on these plurality of reference points, a biometric measurement of the user as well as a measurement of the user's size can be performed. To generate a visual augmented reality, it is preferably the control unit 102 assigns the output of the information or instruction to at least one of the reference points, such that the information or instruction is displayed on the display device 110 in the area of ​​the point on the image and / or on the mirror image corresponding to the reference point.This enables user-friendly guidance, as, for example, the instruction to open the eyes is assigned to the orientation points of the eyes or corners of the eyes and displayed on the image shown by the display device 110 or on the reflection of the semi-transparent mirror. Similarly, the instruction to close the mouth can be displayed on the mouth or corners of the mouth on the image shown by the display device 110 or on the mouth / corner of the mouth in the reflection.

[0050] Preferably, standardized feature points, as listed in Annex C of ISO / IEC 14496-2:2004, are used as reference points. Reference points 3.1 to 3.14 are assigned to the eyes (right eye: pupil 3.6, canthus 3.8 and 3.12; left eye: pupil 3.5, canthus 3.7 and 3.11). Reference points 2.2 to 2.9 and 8.1 to 8.10, respectively, are assigned to the mouth and lips (the corners of the mouth form the pairs 2.4 with 8.3 and 2.5 with 8.4). Reference points 2.1 and 2.12 to 2.14 are assigned to the chin. The nose is assigned the orientation points 9.1 to 9.7 and 9.12 to 9.15, with the tip of the nose being characterized by feature point 9.3. REFERENCE MARK LIST

[0051] 100 Device 101 Camera 102 Control unit 103 Mirror 104 Reflective surface / reflective front 105 Mirror 106 Beam path (camera) 107 Housing 108 Reader 109 Transmitting surface / transmitting back 110 Display device 111 Output unit 112 Display device-free area 113 Communication link (camera) 114 Frame

Claims

1. A device (100) for generating a biometric image of a person's face, comprising a partially transparent mirror (103) whose front surface (104) is reflective and whose rear surface (109), opposite the front surface (104), is transmissive, with at least one display device (110) arranged on the rear side (109) of the partially transparent mirror (103), which is directed toward the transmitting surface of the rear side (109) of the partially transparent mirror (103), and comprising at least one camera (101) for capturing the face of the person, the optical path of which passes through the partially transparent mirror (103) to the face of the person, as well as a control unit (102) and an output unit (111) configured to output information or instructions as visual or auditory user guidance, wherein the control unit (102) and the output unit (111) are configured to output the information or instruction as visual or acoustic user guidance to generate an augmented reality, and that a plurality of display devices (110) are provided, which are arranged relative to one another such that a display-device-free area (112), through which the optical path (106) of the camera (101) passes, is formed between them, wherein an additional deflecting mirror (105) is provided for folding the optical path (106) extending between the camera (101) and the person's face.

2. The device (100) according to claim 1, characterized in that the output unit (111) is configured as the at least one display device (110), and that the information or instruction is displayed as optical user guidance on the display device (110) and is visible on the partially transparent mirror (103).

3. The device (100) according to claim 1 or 2, characterized in that the at least one camera (101) is configured as a three-dimensional camera.

4. The device (100) according to any one of claims 1 to 3, characterized in that the partially transparent mirror (103) is at least partially framed by a frame (114), and that at least one sensor is mounted on or in the frame (114).

5. The device (100) according to any one of claims 1 to 3, characterized in that the partially transparent mirror (103) is framed by a frame (114), and in that lighting elements are mounted on the frame to form a light strip.

6. The device (100) according to claim 5, characterized in that the light strip also extends in a frame-like manner along the entire frame (114).

7. The device (100) according to claim 5 or 6, characterized in that the lighting elements are controllable separately and are controlled to cause the lighting elements of the light strip to illuminate with varying intensity and / or sequentially.

8. The device (100) according to any one of claims 5 to 7, characterized in that the light strip also extends in a frame-like manner and is embedded flush into the frame (114).

9. The device (100) according to any one of claims 1 to 8, characterized in that the partially transparent mirror (103) itself is formed as a touch-sensitive touchscreen.

10. The device (100) according to any one of claims 2 to 9, characterized in that the display device (110) is formed as a touch-sensitive touchscreen, and that the partially transparent mirror (103) is at least partially electrically conductive and is in contact with the display device in such a way that the touchscreen of the display device (110) can be operated through the partially transparent mirror (103).

11. The device (100) according to any one of claims 1 to 10, characterized in that the partially transparent mirror (103), the at least one display device (110), and the at least one camera (101) are arranged in a housing (107), and that the housing (107) is height adjustable.

12. A method for capturing a biometric image of a person's face using a device (100) according to any one of claims 1 to 11, comprising the steps of: A)positioning of a person in front of the reflective front surface (104) of the partially transparent mirror (103) such that the person's face is reflected on the reflective surface (104),B)capturing the person's face via the at least one camera (101) through the partially transparent mirror (103) and generating a first image or a first sequence of images,C)checking whether the image or image sequence generated by the at least one camera (101) meets at least one predetermined requirement,D)if the generated first image or sequence of images deviates from the at least one specified requirement, outputting information or instructions in the output unit (111) as visual or auditory user guidance to create an augmented reality, andE)capturing a further image or sequence of images by the camera (101).

13. The method according to claim 12, characterized in that the control unit (102) performs a position determination of a plurality of reference points in the at least one image or sequence of images relative to the camera (101).

14. The method according to claim 13, characterized in that the control unit (102) assigns the output of the information or instruction to at least one of the feature points such that the information or instruction is displayed on the display device (110) in the area of the point on the image and / or on the mirror image corresponding to the feature point.

15. The method according to claim 13 or 14, characterized in that the predetermined feature points are the mouth and / or the corners of the mouth and / or the eyes and / or the corners of the eyes and / or the chin and / or the nose and / or the tip of the nose.

Images