METHOD AND DEVICE FOR THE BIOMETRIC IDENTIFICATION OF A PERSON AND ACCESS CONTROL DEVICE

Description

[0001] The invention relates to a method and a device for biometrically identifying a person and an access control device. background

[0002] Biometric identification or authentication systems are used to identify individuals, for example, in connection with access control. Biometric data may be captured and evaluated to identify the individual. Examples of biometric data include facial data, fingerprint data, iris data, or data on the course of a hand vein, which are captured and evaluated for personal identification. Frequently changing environmental conditions, such as varying light levels outside buildings, can lead to performance fluctuations in the identification systems.

[0003] Document US 2018 / 0 276 487 A1 provides a method, a system, and a computer-readable recording medium for long-distance personal identification. The method is applicable to a system comprising an image capture device and a depth sensor.

[0004] The document US 2020 / 0 057 847 A1 concerns a contactless recognition method for multiple body parts using multiple biometric data.

[0005] The article: IULA, A. et al. 3D Ultrasound palm vein pattern for biometric recognition. IEEE International Ultrasonics Symposium. October 2012, pages 1 to 4, discloses an ultrasound technique for creating 3D palm vein patterns for biometric recognition purposes.

[0006] In the article: BARBOSA, IB et al. Re-identification with RGB-D Sensors. Computer Vision - ECCV. 2012, pages 433 to 442, a series of soft-biometric 3D indicators are presented for the re-identification of individuals. These indicators are insensitive to optical changes and are captured using RGB-D technology.

[0007] The article: GONZALEZ-SOSA, E. et al. "Feature exploration for biometric recognition using millimeter wave body images." Journal on Image and Video Processing, 2015, Item 30, proposes a biometric recognition system based on information from the human body's silhouette. Summary

[0008] The object of the invention is to provide a method and a device for biometric identification of a person that can be used flexibly under different application conditions and that increases the reliability of correct personal identification. An improved access control device is also to be provided.

[0009] To solve this problem, a method and a device for biometrically identifying a person are provided according to independent claims 1 and 10. Furthermore, an access control device is provided according to independent claim 11. Further embodiments are the subject of dependent subclaims.

[0010] According to one aspect, a method for biometrically identifying a person is provided, which provides the following: providing a biometric detection device; capturing biometric measurement data by means of the detection device, which indicate iris measurement data for an iris of a person to be identified as well as facial measurement data for the face and body measurement data for at least one body portion of the body of the person to be identified, wherein the body measurement data is configured to determine a digital three-dimensional body model for at least the body portion of the person to be identified; providing an evaluation device that receives the biometric measurement data; determining the digital three-dimensional body model for at least the body portion by means of the evaluation device; and determining biometric feature data from the biometric measurement data by means of the evaluation device.The biometric feature data includes: iris data indicating iris features for the person to be identified; facial data indicating facial features for the person to be identified; and body data determined using the digital three-dimensional body model and indicating body features for at least the body portion that are different from the iris features and the facial features. Furthermore, the method provides for providing biometric reference data associated with the person to be identified and indicating iris reference features, facial reference features, and body reference features, and identifying the person to be identified by comparing the biometric feature data with the biometric reference data.

[0011] According to a further aspect, a device for biometrically identifying a person is provided, which device has a biometric detection device which is configured to capture biometric measurement data which, for a person to be identified, indicate iris measurement data for an iris as well as facial measurement data for the face and body measurement data for at least one body portion of the body of the person to be identified, wherein the body measurement data are suitable for determining a digital three-dimensional model for at least the body portion of the person to be identified.The device comprises an evaluation device which receives the biometric measurement data and is configured to: determine the digital three-dimensional body model for at least the body section, determine biometric feature data from the biometric measurement data and provide biometric reference data which are assigned to the person to be identified and indicate iris reference features, facial reference features and body reference features, and identify the person to be identified by comparing biometric feature data with the biometric reference data.The biometric feature data includes the following: iris data indicating iris features for the person to be identified; face data indicating facial features for the person to be identified; and body data determined using the digital three-dimensional body model and indicating body features for at least the body portion that are different from the iris features and the facial features.

[0012] Furthermore, an access control device with a device for biometrically identifying a person is provided, with which admission or entry at an entrance point is controlled depending on the result of the person's identification. For this purpose, identification data is provided in the access control device that indicates the result of the identification of the identifying person. If the person has been identified, admission or entry can be granted. Otherwise, access can be denied. In addition, the access control device can be provided to check whether the identified person is an authorized person to whom admission or entry can be granted. For this purpose, information can be provided in the access control device that indicates one or more authorized persons.If the result of the personal identification was positive, i.e. the person was identified, it can be checked whether this person is one of the authorized persons.

[0013] The biometric measurement data may be captured or provided in whole or in part as digital image data.

[0014] The three-dimensional model for the body can, for example, be determined as a so-called texture model, as is known in connection with digital 3D models, for example in the field of computer graphics.

[0015] The biometric detection device can have an illumination device and a sensor device. When capturing the biometric measurement data, the person to be identified is illuminated with white light and NIR light (NIR - near infrared) by the illumination device, and the sensor device captures white light measurement data associated with the white light illumination and NIR light measurement data associated with the NIR light illumination. For example, it can be provided that for facial recognition, the facial measurement data is captured under white light illumination. When capturing the iris measurement data, NIR light can be used, in particular. When capturing the body measurement data, illumination can be provided with white light and / or NIR light.The lighting device can be designed as an integrated lighting device that includes one or more light sources for white light as well as one or more light sources for NIR light. Additionally, a colored light source for one or more colors (light with a narrower spectral range) can be integrated. Alternatively, separate lighting sources can be provided for the different light types.

[0016] Similarly, the sensor device can be designed as an integrated sensor device configured to capture both the white light measurement data and the NIR light measurement data, in particular as digital image data. Alternatively, separately formed sensor devices can be provided, of which, for example, one or more sensor devices are configured to capture the white light measurement data, and one or more other sensor devices are configured to capture the NIR light measurement data.

[0017] The sensor device may be configured to capture different biometric measurement data with different image resolutions. For example, the iris measurement data may be captured with a higher image resolution than is used for facial and / or body measurement data.

[0018] If personal identification is used in conjunction with an access or entry control system, the lighting device and the sensor device can be arranged in the area of ​​an entrance for which the access or entry control is being carried out. The various elements of the detection device can be arranged at least partially behind the entrance in the direction of the entrance. For example, the lighting device can be arranged behind the entrance to illuminate the person arriving or standing in front of it for biometric data collection. One or more sensor devices can also be arranged in the direction of the entrance in front of or behind the entrance for which the access control is being carried out. In this case, the lighting device and sensor device are arranged in spatial proximity to the entrance itself.

[0019] Different configurations can be provided for the arrangement of light sources of the lighting device. For example, lighting arrangements (lighting arrays) can be arranged alongside a passageway or entrance area through which the person to be identified passes. One or more sensor devices, for example image recording devices, in particular cameras, can be integrated into the lighting arrangement. Alternatively or additionally, the lighting device and / or sensor device for capturing the biometric measurement data can be pivoted or relocated at least partially around the person to be identified in order to capture the biometric measurement data when the person to be identified is standing in a measurement area or room assigned to the detection device. A configuration comparable to a full-body scanner can be provided.

[0020] The white light measurement data can be acquired by a first sensor device, and the NIR light measurement data by a second sensor device. As an alternative to the separate design with the first and second sensor devices, an integrated sensor device can be provided to acquire both the white light measurement data and the NIR light measurement data, for example, an integrated image recording device.

[0021] When capturing the biometric measurement data, the person to be identified can be illuminated with structured light using the lighting device. The structured light can, for example, be formed by spaced-apart light stripes generated on the body surface of the person to be identified by the lighting device. The light stripes can, for example, be arranged in a substantially parallel and / or intersecting configuration. Light patterns, such as light grids or stripe arrangements, can be generated on the person to be identified.

[0022] When capturing biometric data, the person to be identified can be irradiated with terahertz radiation using the illumination device in order to perform a terahertz scan while capturing the biometric data. This can be a full-body scan or a scan of a part of the body. The illumination device and sensor device are configured to perform the terahertz scan. The illumination device can be designed as an integrated illumination device that has one or more radiation sources for the terahertz radiation. Terahertz rays belong to the spectrum of electromagnetic radiation and lie in the frequency range between infrared radiation and microwaves. Clothing is invisible to terahertz detectors. This is because their radiation penetrates materials such as textiles, plastic, paper, and the like almost without difficulty and is barely absorbed.

[0023] When determining the facial data, a digital three-dimensional facial model for the person to be identified can be determined using the facial measurement data, and the facial data can be determined at least partially from the digital three-dimensional facial model. The three-dimensional facial model can, for example, be determined from the associated measurement data, i.e., the facial data, in a manner comparable to the digital three-dimensional model for the body (section). Methods for generating or determining a digital three-dimensional model are known per se in various embodiments. For example, the texture model technique is used. In this case, provision can be made, in particular, to assign a local (point with a certain surroundings) colored appearance in plan view to each detected (image) measurement point or, in triangulation models, to each detected measurement area of ​​the three-dimensional model.

[0024] Distance data can be provided in the evaluation device, indicating the distance between the person to be identified and the sensor device when capturing the biometric measurement data. When determining the body characteristics, the distance data can be used to determine body characteristics characterized by a length measurement. Using the distance data, the digital three-dimensional body model can be generated according to the actual dimensions of the body section, so that body characteristics characterized in particular by a length measurement can then be determined. These include, for example, arm or leg length, hand size, or the length of certain externally visible bones.

[0025] When determining physical characteristics, one or more physical characteristics from the following group can be determined: body mass index, skin tattoos, skin characteristics, and gait characteristics. Regarding skin changes, for example, moles or other specific pigmentations, scars, skin folds, or local hair characteristics can be determined as physical characteristics. Eye and / or skin color can also be used as physical characteristics. Alternatively or additionally, as part of the determination of physical characteristics, it may be planned to determine gait characteristics of the person to be identified from the digital three-dimensional body model, optionally also using a moving digital three-dimensional body model.

[0026] The biometric measurement data can be collected containing body measurement data for at least one body section from the following group: front of the body, back of the body, arms, legs, hands, lower body, and upper body. The digital three-dimensional body model can then be limited to the detected body section. Alternatively, an extended body model with additional body sections and even a full-body model can be determined. In this case, it can be provided to replace body sections of the person to be identified that were not recorded using the body measurement data or that cannot be sufficiently analyzed from the measurement data with sections of a standardized digital three-dimensional body model.

[0027] The following is further provided: providing first biometric reference data that are assigned to the person to be identified and indicate first body reference features; providing second biometric reference data that are assigned to the person to be identified and indicate second body reference features that are different from the first; assigning the first biometric reference data to a first evaluation condition; assigning the second biometric reference data to a second evaluation condition that is different from the first evaluation condition; providing evaluation data in the evaluation device that indicate the first or the second evaluation conditions; and identifying the person to be identified by comparing the biometric feature data with the first biometric reference data, if the evaluation data indicate the first evaluation condition, and with the second biometric reference data,if the evaluation data indicates the second evaluation condition. In this embodiment, different biometric reference data are provided for the person to be identified, namely first and second biometric reference data, which differ. A different evaluation condition is assigned to each of the different biometric reference data. This allows, depending on a specific evaluation condition or situation, certain selected biometric reference data to be used for comparison to identify the person. Thus, it can be provided to determine biometric reference data for the person to be identified,whether the person is wearing winter or summer clothing. Typically, the body surface coverage will be greater in winter clothing. Clothing thickness is also typically greater in winter. As an evaluation condition, the season can then be assigned to different biometric reference data. If the evaluation device has information about the season, the corresponding set of biometric reference data assigned to the current season can be used for comparison. In another example, different reference data can be assigned to different times of day, typically for different ambient lighting conditions (e.g., day and evening). For this purpose, the evaluation device can be provided with a comparison check before identification to determine whether and, if so, which evaluation condition is currently present.in order to then select, depending on this check, a biometric reference data set that is assigned to the evaluation condition found.

[0028] Identification data is generated that indicates the result of the identification of the person to be identified and transmitted to an access control device, whereupon the access control device is controlled according to the result of the identification. In this embodiment, the information about the result of the personal identification is transmitted to the access control device so that it can regulate access to an entrance based on this information, for example, opening a door or gate. The identification data can indicate which types of biometric measurement data were actually used for personal identification—i.e., iris data, facial data, and body data—or just a selection of these.The access control device can be configured to deny access or entry if no body data was used for personal identification, for example, because the recorded body measurement data did not allow for comparison with the body reference characteristics. Access can also be denied, despite the person being identified, if at least two of the different biometric measurement data were not actually used for personal identification—for example, iris data and body data, or facial data and body data.

[0029] The embodiments explained above in connection with the method for biometrically identifying a person can be provided accordingly in connection with the device for biometrically identifying a person and the access control device. Description of implementation examples

[0030] Further embodiments are explained below with reference to the figures of a drawing. Herein: Fig. 1 shows a schematic representation of an entrance area with a wall section and a wall passage, as well as an associated detection device; Fig. 2 shows a schematic representation of another entrance area with an associated detection device; Fig. 3 shows a schematic block diagram of an access control device; and Fig. 4 shows a schematic block diagram of a method for biometrically identifying a person.

[0031] Fig. 1shows a schematic representation of an entrance area 1 in plan view, in which a passage (entrance) 4 for a person 5 to be identified is formed between opposing wall sections 2, 3. On one side of the two wall sections 2, 3, an illumination device 6 with a first and second light source 6a, 6b is arranged. On the opposite side of the two wall sections 2, 3, a sensor device 7 with a first and a second image recording device 7a, 7b is provided. With the aid of the illumination device 6, the person 5 to be identified, who enters the entrance area 1, can be illuminated, in particular with white light and / or NIR light, in order to then, with the aid of the sensor device 7, record biometric measurement data for the person, namely iris data for an iris of the person 5, facial data for the face of the person 5, and body data for at least one body section of the person 5.The collected biometric measurement data can then be evaluated for personal identification, which is explained further below.

[0032] Fig. 2 shows a schematic representation of another entrance area 10 from the front, in which light sources 11a, ..., 11d of a lighting device 11 are arranged on opposite sides to illuminate the person 5 to be identified, who enters the entrance area 10, with one or more types of light. A sensor device 12 is formed on both sides with camera devices 12a, 12b to capture biometric measurement data for the person 5 to be identified. In the illustrated embodiment, the lighting device 11 and the sensor device 12 are formed in opposing column arrangements 13a, 13b.

[0033] Fig. 3shows a schematic representation of an access control device 20, which controls an access or entry device 21, for example a door or a gate. The access control device 20 is connected to an evaluation device 22, which in turn is connected to a detection device 23. The detection device 23 has an illumination device and a sensor device for capturing biometric measurement data for the person 5 to be identified, in particular one or more image recording devices, as described with reference to the Fig. 1 and 2 as explained above using examples.

[0034] With reference to Fig. 4A method for biometrically identifying the person 5 to be identified is explained below. According to step 30, biometric measurement data for the person 5 to be identified is acquired by means of the detection device 23. The biometric measurement data comprises the following various biometric data: iris data for an iris of the person 5 to be identified; facial data for the face of the person 5 to be identified; and body measurement data for at least one body section of the person 5 to be identified. The body measurement data are suitable for use in generating a digital three-dimensional body model for at least the body section. For this purpose, it can be provided that the body measurement data are acquired by means of a plurality of image recording devices that are spatially distributed around the person 5 to be identified when acquiring the biometric measurement data.Alternatively or additionally, it can be provided that light structures are generated on the person 5 to be identified with the aid of the lighting device in order to record suitable body measurement data.

[0035] In step 31, the digital three-dimensional body model for at least one body section or the entire body of the person 5 to be identified is determined from the body measurement data. Techniques suitable for generating a digital three-dimensional body model from such body measurement data are known in various embodiments. If individual areas of the body of the person 5 to be identified are not captured using the body measurement data, they can be replaced by standardized body surfaces or sections when creating the digital three-dimensional body model. For missing areas for which no measurement data is available, a gray surface without a specific texture could be inserted into a visualization, so that these areas are "masked."

[0036] In step 32, body characteristics for the person 5 to be identified are determined from the digital three-dimensional body model. These can include body characteristics that are characterized by a length measurement, for example leg length, arm length, or hand size. Such a dimensional determination of body characteristics is possible if the digital three-dimensional body model was determined according to the actual body measurements of the person 5 to be identified. For this purpose, distance information is used that indicates the distance between the image recording devices 7a, 7b, 12a, 12b used to capture the biometric measurement data and the person 5 to be identified. Alternatively or additionally, body characteristics that are not determined by a length measurement can also be used; for example, a so-called body mass index, eye color, and / or skin color can be determined for the person to be identified.

[0037] In step 33, iris features for the person 5 to be identified are determined from the biometric measurement data. Furthermore, facial features are determined from the biometric measurement data in step 34.

[0038] Biometric reference data is then provided, which is assigned to the person 5 to be identified and indicates iris reference features, facial reference features, and body reference features. To identify the person 5, the biometric reference data is compared with the biometric feature data in step 36. In this case, it can be provided that the body data and at least the iris data or at least the facial data are actually used for the person's identification in every case, for example, if the iris data or the facial data do not have sufficient quality for comparison with the biometric reference data. Evaluating all three types of biometric data—namely, iris data, facial data, and body data—improves the reliability of the person's identification process.

[0039] Optionally, it may be provided to transmit the information about the successful personal identification to the access control device 20, which can then control the access or entry depending on the personal identification.

[0040] In connection with the body measurement data, it can be provided that, for example, only the front or the back of the body of the person 5 to be identified is recorded. It can also be provided that body measurement data are recorded only for the upper body, from the front and back.

[0041] For access control, it can be assumed that people want to enter somewhere. They will therefore regularly pass through a door or similar structural element. This leads them to regularly behave in such a way that they look in the direction of the corridor and, in a foreseeable sequence, find themselves in predictable, closely spaced locations. This is exploited. The following data should be recorded: facial data, iris data, and body shape. Ideally, the body shape encompasses the entire body.

[0042] In the area of ​​the real or virtual door, suitable light or lighting sources are first installed (see Fig. 1 and 2 and explanations above), which, for example, cover the entire visible spectrum (white light) and the near-infrared range (NIR) and the passage point of person 5 through the door and a certain area in front of and behind it (for example, about 50 cm) (spatially).

[0043] Data from the eyes (with high resolution, NIR and white light), face, and body surface (with good resolution, NIR and white light) of person 5 located in this area are to be recorded using (in most cases several) cameras. The idea is to create a digital three-dimensional model, for example by applying structured light or by using at least two cameras. It is advisable to use a different (higher resolution) camera for the area requiring particularly high resolution (eye region) than for the rest of the body. It is also advisable to use different cameras for white light and NIR; with suitable lenses and interchangeable filters, solutions for NIR and white light with a single camera are also possible.

[0044] The acquired data is used to create a digital 3D model of the body, similar to that used, for example, with vertex representations of the face in 3D facial recognition. Regions for which no data is available can be omitted. This 3D model is enriched with appropriate color information, where color here can include both the white light and NIR response. This creates the data set used for comparisons. For both the facial and iris information, standard methods can be used to generate conventional two-dimensional data sets of these characteristics. If the body's position relative to the cameras is known, all visible distances can be measured. These measured values ​​contribute to a so-called feature set of the biometric data to be compared.Data derived from facial information using common methods, such as body mass index, can be checked for consistency with the overall body appearance.

[0045] At times when the person to be identified is wearing thicker clothing, the recorded body data may differ from those at times when they were wearing lighter clothing. This is taken into account in the comparison model.

[0046] The digital three-dimensional body model can use measured lengths for individual body parts to scale. In cases where large portions of the body surface are visible, such as at the entrance to a swimming pool, a standardized body model can be used alternatively, in which all body surfaces are scaled to a standard body. This approach loses information about the body shape, but allows for easy comparison of surface information and its structure between different individuals, for example, skin folds, freckles, or moles.

[0047] The features disclosed in the above description, the claims and the drawings may be important for the realization of the various embodiments both individually and in any combination.

Claims

1. A method for biometrically identifying a person (5), comprising: - providing a biometric detection device (23); - acquiring biometric measurement data by means of the detection device (23), which indicate iris measurement data for an iris of a person (5) to be identified as well as face measurement data for the face and body measurement data for at least one body section of the body of the person (5) to be identified, wherein the body measurement data are suitable for determining a digital three-dimensional body model for at least the body section of the person to be identified; - providing an evaluation device (22), which receives the biometric measurement data; - determining the digital three-dimensional body model for at least the body section by means of the evaluation device (22); - determining biometric feature data from the biometric measurement data by means of the evaluation device (22), which comprise: - iris data indicating iris features for the person (5) to be identified; - face data indicating face features for the person (5) to be identified; and - body data which are determined by means of the digital three-dimensional body model and indicate body features for at least the body section which are different from the iris features and the face features; - providing biometric reference data which are associated with the person (5) to be identified and indicate iris reference features, face reference features and body reference features, wherein first biometric reference data associated with the person (5) are provided, which indicate first body reference features, and second biometric reference data associated with the person (5) are provided, which indicate second body reference features which are different from the first body reference features; - associating the first biometric reference data with a first evaluation condition and associating the second biometric reference data with a second evaluation condition which is different from the first evaluation condition; - providing evaluation data in the evaluation device, which indicate the first or the second evaluation condition; and - identifying the person (5) to be identified by means of comparing the biometric feature data with the biometric reference data, wherein the biometric feature data are compared with the first biometric reference data if the evaluation data indicate the first evaluation condition, and are compared with the second biometric reference data if the evaluation data indicate the second evaluation condition, wherein identification data indicating a result of the identification of the person (5) to be identified are generated and transmitted to an access control device (20), whereupon the access control device (20) is controlled according to the result of the identification.

2. The method according to claim 1, characterized in that the biometric detection device (23) comprises an illumination device (6; 11) and a sensor device (7; 12), wherein, during the acquisition of the biometric measurement data, the person (5) to be identified is illuminated with white light and NIR light by means of the illumination device (6; 11) and white light measurement data associated with the illumination with white light and NIR light measurement data associated with the illumination with NIR light are acquired by means of the sensor device (7; 12).

3. The method according to claim 2, characterized in that the white light measurement data are acquired by means of a first sensor device and the NIR light measurement data are acquired by means of a second sensor device.

4. The method according to at least one of the preceding claims, characterized in that, during the acquisition of the biometric measurement data, the person (5) to be identified is illuminated with structured light by means of the illumination device (6; 11).

5. The method according to at least one of the preceding claims, characterized in that, during the acquisition of the biometric measurement data, the person (5) to be identified is irradiated with terahertz radiation by means of the illumination device (6; 11) in order to carry out a terahertz scan during the acquisition of the biometric measurement data.

6. The method according to at least one of the preceding claims, characterized in that, during the determination of the face data, a digital three-dimensional face model for the person (5) to be identified is determined using the face measurement data and the face data are determined at least partially from the digital three-dimensional face model.

7. The method according to at least one of the preceding claims, characterized in that - distance data are provided in the evaluation device (22), which indicate a distance between the person (5) to be identified and the sensor device (7; 12) during the acquisition of the biometric measurement data; and - body features characterized by means of a length measure are determined during the determination of the body features using the distance data.

8. The method according to at least one of the preceding claims, characterized in that, during the determination of the body features, one or more body features from the following group are determined: body mass index, skin tattoos, skin changes and gait properties.

9. The method according to at least one of the preceding claims, characterized in that biometric measurement data including body measurement data for at least one body section from the following group are acquired: front side of the body, back side of the body, arms, legs, hands, lower body and upper body.

10. An apparatus for biometrically identifying a person (5), comprising: - a biometric detection device which is configured to acquire biometric measurement data which indicate, for a person (5) to be identified, iris measurement data for an iris as well as face measurement data for the face and body measurement data for at least one body section of the body of the person (5) to be identified, wherein the body measurement data are configured to determine a digital three-dimensional model for at least the body section of the person (5) to be identified; - an evaluation device, which receives the biometric measurement data and is configured to - determine the digital three-dimensional body model for at least the body section; - determine biometric feature data from the biometric measurement data, which comprise: - iris data indicating iris features for the person (5) to be identified; - face data indicating face features for the person (5) to be identified; and - body data which are determined by means of the digital three-dimensional body model and indicate body features for at least the body section which are different from the iris features and the face features; - provide biometric reference data which are associated with the person (5) to be identified and indicate iris reference features, face reference features and body reference features, wherein first biometric reference data associated with the person (5) are provided, which indicate first body reference features, and second biometric reference data associated with the person (5) are provided, which indicate second body reference features which are different from the first body reference features; - associate the first biometric reference data with a first evaluation condition and the second biometric reference data with a second evaluation condition which is different from the first evaluation condition; - provide evaluation data in the evaluation device, which indicate the first or the second evaluation condition; and - identify the person (5) to be identified by means of comparing the biometric feature data with the biometric reference data, wherein the biometric feature data are compared with the first biometric reference data if the evaluation data indicate the first evaluation condition, and are compared with the second biometric reference data if the evaluation data indicate the second evaluation condition, - generate and transmit identification data indicating a result of the identification of the person (5) to be identified to an access control device (20), whereupon the access control device (20) is controllable according to the result of the identification.

11. An access control device (20), comprising an apparatus for biometrically identifying a person (5) according to claim 10.