METHOD FOR IDENTIFYING A PERSON BY MEANS OF A BIOMETRIC FACIAL FEDERATION USING A MOBILE DEVICE AND MOBILE DEVICE

Description

[0001] The invention relates to a method for identifying a person based on a biometric facial feature using a mobile device and mobile device. background

[0002] It is known to identify individuals based on biometric facial features. In one implementation, facial features are derived from two-dimensional facial images and compared with reference or comparison data that provide facial comparison features in order to identify the person.

[0003] Facial recognition, in this context, generally refers to the analysis of the expression of visible features in the frontal region of the head, determined by the geometric arrangement and texture properties of the head's surface. So-called landmarks, which relate to characteristic positions in the frontal region, are used to characterize faces. Landmarks as such have been studied and are well-known (see, for example, Luximon et al.: "The Chinese face: A 3D anthropometric analysis", Proceedings of the TMCE 2010, April 12-16, 2010, Ancona, Italy).

[0004] Providing suitable two-dimensional facial images or representations is a frequent problem. When personalizing ID documents (ID identification), for example at a citizens' registration office or residents' registration office, two-dimensional facial images are provided as analog printouts or captured with a 2D camera. It is often difficult to provide distortion-free two-dimensional images that are also suitable for determining facial features from the captured image data for biometric recognition or identification of the person.

[0005] To achieve the desired quality of two-dimensional facial images, image capture parameters are defined, such as the distance between the camera and the person, lighting and background, as well as camera characteristics like resolution, distortion, and color fidelity. There are also specifications for other parameters relevant to recognition, such as pose, facial expression, open eyes, and the absence of any obstruction of the eyes by hairstyle, glasses, or headwear. Therefore, providing such two-dimensional images suitable for biometric facial recognition is a complex and costly process.

[0006] Document DE 10 2011 004 787 A1 discloses an ID document for identifying a person, comprising an image of a first view of the person and model information that enables the digital capture of a second view of the person. To identify the person, the second view is displayed via a screen, allowing a controller to compare the first view on the identification document with the second view on the screen. Automatic comparison of the two views may also be possible.

[0007] In BUSCH C ET AL: "Towards a more secure border control with 3D face recognition", PROCEEDINGS OF THE 5TH NORSK INFOMASJONS SIKKERHETS KONFERANSE (NISK); November 19, 2012, the comparison of a 3D face scan with a 2D reference image is disclosed. The 3D face scans include biometric data. A capture device is used to record personal data. Furthermore, the analysis of a 3D model and the derivation of features, as well as their comparison with reference data, are disclosed; in this process, facial landmarks are located and used as reference marks. A 3D Face_ID Card is also disclosed. The aforementioned means and measures serve to identify individuals. Summary

[0008] The object of the invention is to provide a method for identifying a person using a biometric facial feature with the aid of a mobile device, as well as a mobile device that enables secure person identification that can be used in various application situations.

[0009] To solve this problem, a method for identifying a person based on a biometric facial feature using a mobile device according to independent claim 1, and a mobile device according to independent claim 14, have been provided. Embodiments are the subject of dependent claims.

[0010] According to one aspect, a method for identifying a person using a biometric facial feature with the aid of a mobile device is created, in which an ID document with a storage for electronic data is provided, wherein, to personalize the ID document for a person to whom the ID document is assigned, at least one of the following data sets is stored in the storage: reference value data which indicates a reference value for a geometric relationship for at least two landmarks of facial recognition in a three-dimensional reference model for the person's face; landmark data which indicates landmarks for facial recognition in the three-dimensional reference model and from which the reference value data can be determined;and model data representing the three-dimensional reference model from which the reference value data can be determined. Furthermore, the following is provided: reading at least one of the data sets from the memory of the ID document and providing the reference value data to at least one processor of a mobile device; capturing current image data of the person's face with an image capture device of the mobile device; determining a current three-dimensional model of the person's face using the current image data; determining comparison data from the current three-dimensional model, wherein the comparison data indicates a comparison value for the geometric parameter for the at least two facial recognition landmarks determined in the current three-dimensional model of the face, and providing the comparison data to the at least one processor of the mobile device;and identifying the person when, using at least one processor of the mobile device, it is determined that the comparison value is equal to the reference value within a specified error tolerance.

[0011] According to another aspect, a mobile device for identifying a person based on a biometric facial feature is created, which includes at least one processor, a data communication device, and an image capture device. The mobile device is configured to: read at least one data set from the memory of an ID document, wherein the at least one data set for personalizing the ID document for a person to whom the ID document is assigned is stored in the memory and comprises at least one of the following data sets: reference value data indicating a reference value for a geometric relationship for at least two facial recognition landmarks in a three-dimensional reference model for the person's face; landmark data indicating facial recognition landmarks in the three-dimensional reference model from which the reference value data can be determined; and model data.which represent the three-dimensional reference model and from which the reference value data can be determined; providing the reference value data in the at least one processor; capturing current image data for the person's face with the image capture device; determining a current three-dimensional model for the person's face using the image data; determining comparison data from the current three-dimensional model, wherein the comparison data indicates a comparison value for the geometric relationship between the at least two facial recognition landmarks in the current three-dimensional model for the face, and providing the comparison data in the at least one processor; and identifying the person when, using the at least one processor, it is determined that the comparison value is equal to the reference value within a specified error tolerance.

[0012] It is intended that the electronic data storage of the ID document will contain one or more data sets that enable the determination of reference value data or reference data for a biometric facial feature, or that themselves comprise such data. Storing the one or more data sets is part of a personalization procedure for the ID document. The reference value data indicates a reference value for at least one geometric parameter or a geometric relationship relating to at least two landmarks of facial recognition, in particular a geometric parameter that indicates properties for one or more connecting lines between the at least two landmarks.The reference value data was determined from a three-dimensional reference model or a three-dimensional reference reconstruction of the person's face. The three-dimensional reference model is configured and detailed enough for the analysis or determination of features (landmarks) for biometric facial recognition. If the ID document contains (only) the landmark data, which is itself determined from the three-dimensional reference model, the reference value relating to the geometric relationship (reference value data) can be determined from this data. For example, a distance value between two landmarks is determined, as indicated by the landmark data, perhaps based on their respective position data in the three-dimensional reference model.

[0013] The three-dimensional reference model captures at least the visible area of ​​the frontal head, but can also be a model of the entire head. For facial recognition, landmarks can be defined on the three-dimensional reference model; that is, the positions of selected characteristic points on the reference model. Then, a geometric parameter is determined for two or more landmarks, for example, the distance between them.

[0014] The identification process involves comparing reference value data stored in the ID document's memory, or determined from the data set(s) stored there, with comparison data to identify the person. This comparison is valid only if the reference value displayed by the reference data and a comparison value displayed by the comparison data for the geometric parameter relating to at least two landmarks match within a tolerance range. The tolerance range can be chosen depending on the application, as is common practice when comparing a reference value with a comparison value determined from a current measurement.

[0015] The comparison data is determined from a current three-dimensional model or a current three-dimensional reconstruction of the person's face, which is determined using image data for the purpose of current personal identification, whereby the image data was recorded prior to its evaluation for determining the comparison data using an image recording device of a mobile device for the person to be identified, who is currently presenting or submitting the ID document for identification, in particular at the time of or as part of the personal identification.

[0016] The process, using a mobile device and an ID document with the data stored on it, enables efficient and secure personal identification in a wide variety of applications, such as entry control into a country or access control for company premises. One advantage is that using a value (reference value, comparison value) determined from three-dimensional models or reconstructions is less error-prone than measurements taken in facial recognition based on two-dimensional images. The comparison value, determined using the comparison data, is calculated during personal identification based on the image data currently being captured of the person to whom the ID document belongs and who is presenting or using the ID document for identification purposes.

[0017] The ID document can be, in particular, an identification, security, or other document, especially an official document, for example, a paper-based and / or plastic-based document, such as an electronic identification document, in particular a passport, identity card, visa, driver's license, vehicle registration certificate, vehicle title, health insurance card, or company ID card, or other ID document such as proof of authorization. In one form, the ID document can be a chip card.

[0018] Storing the data(s) in the ID document's memory can occur before the memory (e.g., a chip or memory module) is integrated into the document body. Alternatively, the data can be stored after the memory has been incorporated into the document body.

[0019] At least one of the data sets can be read from the ID document's memory using a data communication device on the mobile device connected to at least one processor. The mobile device's data communication device is configured to read the data set(s) directly from the ID document's electronic data memory, either via contactless and / or contact-based data transmission. Thus, the mobile device itself can be used to load the data set(s) for personal identification. For this purpose, the mobile device can, for example, have a chip card reader that enables contactless and / or contact-based memory readout.

[0020] The current three-dimensional model can be determined or calculated locally in the mobile device using at least one processor. For this purpose, the mobile device must have suitable software applications. Alternatively, it may be possible to calculate at least part of the current three-dimensional model of the face in an external data processing unit with which the mobile device can exchange data, either wirelessly and / or via a wired data connection. This allows for the use of external computing power to supplement the processor or computing capacity available in the mobile device in order to determine the current three-dimensional model.In this case, it may be provided that the current image data recorded with the help of the mobile device are transmitted to the external data processing facility and that the electronic data for the current three-dimensional model are transmitted in whole or in part from the external data processing facility to the mobile device.

[0021] The comparison data can be determined locally in the mobile device using at least one processor. The explanations regarding the optional use of an external data processing device apply accordingly. The comparison data to be determined or derived from the current three-dimensional model can be determined, at least partially, using the external data processing device and then transmitted from there to the mobile device.

[0022] The at least two landmarks for facial recognition can include at least one of the following landmarks: position of the tip of the nose, position of the eye socket of the left eye, position of the eye socket of the right eye, position of the right ear, and position of the left ear.

[0023] In principle, any landmarks can be selected to define one or more geometric parameters or relationships, particularly concerning connecting lines between the selected landmarks. The reference value data for these parameters is then stored in the ID document's memory. Landmarks are preferred that are less susceptible to errors caused by varying measurement conditions during image data acquisition using the mobile device and the subsequent calculation of the current three-dimensional model.When selecting preferred landmarks, one or more of the following criteria can be applied to landmark properties: low dynamics due to the landmark's facial expressions; high "singularity" of the landmark, for example, the tip of the nose, the depression near the eyes next to the nose, or the like; local minima or maxima; high entropy of the landmark across all people, i.e., a lot of information per pixel; high variance per mm of length and / or per angle, for example, triangular shape.

[0024] The geometric relationship for the at least two landmarks used for facial recognition can specify a relationship from the following group: distance between two landmarks; angle between two lines, each corresponding to a connecting line between two landmarks; and geometric shape such as triangle, circle, or quadrilateral. The geometric parameter or relationship can indicate information regarding the geometric shape of connecting lines between landmarks, for example, a triangle or quadrilateral. The geometric relationship can, for instance, relate to the distance between the left eye socket and the right eye socket. Alternatively or additionally, the reference data can indicate the distance between the left eye socket and the tip of the nose, as well as between the right eye socket and the tip of the nose.

[0025] A mobile device from the following group can be used: mobile phone, laptop computer, tablet computer, and handheld scanner. The mobile device can be configured with one or more locally installed software applications to determine the current three-dimensional model of the person's face, enabling the model to be calculated locally on the mobile device to the extent necessary for determining the comparison data. Alternatively or additionally, an external data processing device can be used to perform the calculations.

[0026] The memory may contain additional reference value data for another biometric facial feature, indicating a further reference value for a further geometric relationship between at least two facial recognition landmarks in the three-dimensional reference model of the person's face. These landmarks may be the same or different ones. Alternatively or additionally, further landmark data may be stored on the ID document, enabling the determination of this further reference value.

[0027] It can be provided that the person is only identified if, for several reference values, each relating to a geometric parameter between landmarks of the facial recognition system, the reference value and its associated comparison value match within the same or different tolerance ranges. Alternatively, it can be provided that the match for only one reference value is sufficient for identifying the person. The invention provides that the several reference values ​​are weighted reference values ​​and are checked sequentially according to a weighted order, the check sequence being terminated if a match with the respective associated comparison value from the current three-dimensional model is found within the respective tolerance range for one or more of the reference values.

[0028] At least one data set can be stored in the ID document's memory in accordance with the specifications for the logical data structure for storing biometric information in machine-readable travel documents according to ICAO Doc 9303 (7th edition, 2015 or later). Part 10 of the 7th edition of ICAO Doc 9303 specifies the requirements for the logical data structure for storing biometric information. It defines a standardized logical data structure with data groups 1 to 16, as specified in this edition. A hash value is stored for each data group in a so-called Document Security Object (DSO). It may be possible to store at least one data set in one of the data groups in the ID document's memory that is intended for future use according to ICAO Doc 9303, i.e., data that is not already occupied by other data according to the current edition of ICAO Doc 9303.

[0029] The authenticity and integrity of at least one data set can be protected by means of an electronic signature for passive authentication for machine-readable travel documents in accordance with ICAO Doc 9303 (7th edition, 2015 or later). Part 12 of ICAO Doc 9303, 7th edition (2015), specifies a public key infrastructure for machine-readable travel documents. The authenticity and integrity of the data stored in the machine-readable travel document is protected by means of passive authentication. This security mechanism is based on digital signatures and includes the elements of the public key infrastructure (see "3. Rolls and Responsibilities" in Part 12 of ICAO Doc 9303, 7th edition, 2015).

[0030] When identifying the person, additional personal data can optionally be read and checked using the mobile device, for example the data intended for machine-readable travel documents according to ICAO Doc 9303.

[0031] The reference value data can be determined using the model data when identifying the person. Determining the reference value data can be performed locally on the mobile device using at least one processor. For this purpose, the mobile device must have suitable software applications. Alternatively, it can be provided that the reference value data is calculated, at least partially, in the external data processing facility with which the mobile device can exchange data. This allows for the use of external computing power to supplement the processor or computing capacity available in the mobile device in order to determine the reference value data. For this purpose, the model data can be transmitted from the mobile device to the external data processing facility.

[0032] The reference value data can be determined using landmark data when identifying the person. This determination can also be performed locally on the mobile device using at least one processor. For this purpose, the mobile device must be equipped with suitable software applications. Alternatively, it can be provided that the reference value data is calculated, at least partially, in the external data processing facility. This allows for the use of external computing power to supplement the processor or computing capacity available in the mobile device in order to determine the reference value data. In this case, it can be provided that the landmark data is transmitted from the mobile device to the external data processing facility for the purpose of determining the reference value data.

[0033] If at least one data set read from the ID document does not include the reference value data, the processor of the mobile device and / or the external data processing unit provides information specifying which landmarks and / or geometric relationships between landmarks should be used to identify the person, optionally depending on the application. The reference value data is then determined according to these specifications. For this purpose, corresponding information or verification data can be stored in the memory of the ID document, the mobile device, and / or the external data processing unit. This data can be evaluated during identification and thus influence the identification process.

[0034] A software application can be implemented on the mobile device to assist the user in capturing current image data of the person. This application checks, for example, whether the captured image data is (already) suitable for calculating the current three-dimensional model. If not, the user can be prompted to take further images until a sufficient amount or basis of data for model calculation is available. To guide the user during the image capture process, a suitable user menu can be provided on the mobile device's display via the software application.

[0035] The identification process allows for the recording of current image data for the person's face when the person presents the ID document for identification.

[0036] The configurations described above in connection with the procedure for identifying a person may be provided accordingly in connection with the mobile device for identifying the person based on the biometric facial feature. Description of exemplary implementations

[0037] Further examples of implementation are explained below with reference to figures in a drawing. These show: Fig. 1 a schematic representation of an arrangement for identifying a person using a biometric facial feature; Fig. 2 a schematic representation to illustrate a method for identifying a person using a biometric facial feature; and Fig. 3 a schematic representation of part of a three-dimensional model for the face of a person to be identified.

[0038] Fig. 1Figure 1 shows a schematic representation of an arrangement for identifying a person 1 using one or more biometric facial features. Fig. 2 shows a schematic representation of a flowchart to describe a procedure for identifying person 1.

[0039] In step 20, reference value data for person 1 are stored in an electronic memory 2 of an ID document 3. The ID document 2 can be, in particular, an identification, value or security document, especially an official document, for example, a paper-based and / or plastic-based document, such as an electronic identification document, in particular a passport, identity card, visa, driver's license, vehicle registration certificate, vehicle title, health card or company ID card or other ID document such as proof of authorization.

[0040] The reference value data can be stored in memory 2 of ID document 3 before memory 2, for example a chip or a memory module, is integrated into a document body of ID document 3. Alternatively, the reference value data can be stored after memory 2 has been included in the document body.

[0041] The reference value data relating to a biometric facial feature indicates at least one reference value for one or more geometric parameters / geometric relationships for at least two landmarks used for facial recognition. For example, this could include the distance between the positions of two landmarks and / or an angle between connecting lines between the respective positions of two landmarks. The reference value was determined from a three-dimensional reference model or a three-dimensional reference reconstruction of Person 1's face. Therefore, the three-dimensional reference model for Person 1's face was calculated or determined before or during the personalization of ID document 3.During personalization, one or more of the following data sets or records are stored in memory 2: reference value data, which indicates a reference value for a geometric relationship between at least two facial recognition landmarks in the three-dimensional reference model for the person's face; landmark data, which indicates facial recognition landmarks in the three-dimensional reference model and from which the reference value data can be determined; and model data, which represents the three-dimensional reference model and from which the reference value data can be determined. Depending on which data set(s) are stored during personalization, the three-dimensional reference model, or from it the landmark data and / or the reference value data, must be determined for the purpose of personalization.

[0042] The reference value is a property for facial recognition that is derived from the three-dimensional reference model and thus represents a characteristic of the three-dimensional reference model for the person's face.

[0043] In the process for identifying person 1, in step 20 at least one of the previously stored data sets is read from the memory 2 of the ID document 3 using a data communication device 4 of a mobile device 5 and made available in a processor 6 of the mobile device 5. Here, the memory 2 of the ID document 3 is read directly via an interface of the ID document 3 using the data communication device 4 of the mobile device 5, in particular by means of wireless data communication, as is known, for example, for reading and writing electronic data in connection with so-called chip cards. The data communication device 4 of the mobile device 5 can, for example, have a card reader or be connected to such an external card reader.

[0044] If at least one of the read data sets does not include the reference value data, but rather the landmark data and / or the model data, the reference value data indicating the reference value will be determined, either locally by means of the processor and / or an external data processing device.

[0045] In step 21, current image data of the face of the person 1 to be identified is captured using an image capture device 7, which is integrated into the mobile device 5 and consists, for example, of a camera. Based on this current image data, a current three-dimensional model or reconstruction for facial recognition of person 1 is determined in step 22, for example, as a head model. The calculation of the current three-dimensional model can be performed partially or entirely locally using the processor 6 of the mobile device 5, which may also have multiple processors and one or more software applications to perform the necessary calculations.Alternatively or additionally, it may be possible to perform the determination of the current three-dimensional model using the current image data, either wholly or partially, in the external data processing unit 8, which, in the example shown, is connected to a database 9 and configured to exchange data with the mobile device 5, either via a wired or wireless data connection. In this way, additional processor or computing capacity can be used selectively for the various calculation steps.

[0046] In step 23, comparison data is determined from the current three-dimensional model of person 1's face. This comparison data represents a reference value for the geometric parameter / relationship for at least two landmarks used for facial recognition and is derived from or determined within the current three-dimensional model. The comparison data, like the reference value data, is provided to the processor 6 of the mobile device 5. In step 24, person 1 is identified when the processor 6 of the mobile device 5 determines that the comparison value is equal to the reference value within a predefined error tolerance. In this way, person 1 can be efficiently and reliably identified in a wide variety of application scenarios using the mobile device 5 and the ID document 3.The identification process is based on the reference value data and the comparison data, each determined from a three-dimensional representation or reconstruction of the face of person 1.

[0047] If at least one of the retrieved data sets does not include the reference value data, information is provided in the processor and / or the external data processing unit 8 regarding which landmarks and / or which geometric relationships between landmarks should be used to identify the person, so that the reference value data can then be determined accordingly. For this purpose, corresponding information or verification data may be stored in memory 2 of the ID document 3, which thus also influences the identification process.

[0048] Fig. 3Figure 1 shows a schematic representation of an example of a front view of a three-dimensional model of a person. A multitude of landmarks 31 are drawn in the facial area 30, each serving as a landmark for facial recognition. As an example of geometric parameters relating to selected landmarks, connecting lines 32 between the position of the left eye socket and the position of the right eye socket, as well as the position of the tip of the nose, are shown. One or more of these geometric parameters, for example, the distance and / or angle between the connecting lines 32, can be used to be stored as reference value data in the ID document 3 and then used for person identification.Due to the large number of landmarks 31, different landmarks can be selected in order to determine one or more geometric parameters / relationships for them, for example the distance between landmarks and / or an angle between connecting lines that connect two or more landmarks.

[0049] Storing at least one set of data in ID document 3 can be done according to the specifications of ICAO Doc 9303 (7th edition, 2015 or later), especially if ID document 3 is a machine-readable travel document. However, it may also be possible to apply these standardized specifications to a different ID document that is not a travel document but is machine-readable.

[0050] The features disclosed in the foregoing description, the claims and the drawing can be important for the realization of the various embodiments, both individually and in any combination. Reference symbol list

[0051] 1 Person 2 Storage 3 ID document 4 Data communication device 5 Mobile device 6 Processor 7 Image capture device 8 External data processing device 9 Database 20 ... 24 Procedure steps 30 Facial area 31 Landmarks 32 Connecting lines

Claims

1. Method for identifying a person on the basis of a biometric facial feature with the aid of a mobile device, comprising: - providing an ID document (3) with a memory (2) for electronic data, wherein, in order to personalize the ID document (3) for a person to whom the ID document (3) is assigned, at least one of the following data sets is stored in the memory (2): - reference value data which indicate a plurality of reference values for geometric relationships for at least two landmarks of the facial recognition in a three-dimensional reference model for the face of the person, - landmark data which indicate landmarks for facial recognition in the three-dimensional reference model and from which the reference value data can be determined, and - model data which represent the three-dimensional reference model and from which the reference value data can be determined; - reading at least one of the data sets from the memory (2) of the ID document (3) and providing the reference value data in at least one processor of a mobile device (5); - recording current image data for the person's face using an image recording device (7) of the mobile device (5); - determining a current three-dimensional model for the person's face using the current image data; - determining comparison data from the current three-dimensional model for the person's face, wherein the comparison data indicate a plurality of comparison values for geometric parameters for the at least two landmarks of the facial recognition which are determined in the current three-dimensional model, and providing the comparison data in the at least one processor (6) of the mobile device (5); and - identifying the person if it is determined by means of the at least one processor (6) of the mobile device (5) that at least one of the comparison values is equal to a corresponding reference value within a predefined error tolerance; characterized in that the reference values are weighted reference values and, when identifying the person, are checked in accordance with a weighted sequence one after the other, wherein the checking sequence is aborted if a correspondence with a respectively assigned comparison value within a respective tolerance range is established for one or more of the reference values.

2. Method according to Claim 1, characterized in that the at least one of the data sets is read from the memory (2) of the ID document (3) with the aid of a data communication device of the mobile device (5) which is connected to the at least one processor (6).

3. Method according to Claim 1 or 2, characterized in that the current three-dimensional model is determined locally in the mobile device (5) by means of the at least one processor (6).

4. Method according to at least one of the preceding claims, characterized in that the comparison data are determined locally in the mobile device (5) by means of the at least one processor (6).

5. Method according to at least one of the preceding claims, characterized in that the at least two landmarks of the facial recognition comprise at least one of the following landmarks: position of the tip of the nose, position of the eye depression of the left eye, position of the eye depression of the right eye, position of the right ear and position of the left ear.

6. Method according to at least one of the preceding claims, characterized in that the geometric relationship for the at least two landmarks of the facial recognition indicates a relationship from the following group: distance between two landmarks; angle between two straight lines which each correspond to a connecting line between two landmarks; and geometric shape such as triangle, circle or square.

7. Method according to at least one of the preceding claims, characterized in that a device from the following group is used as mobile device (5): mobile telephone, laptop computer, tablet computer and hand scanner.

8. Method according to at least one of the preceding claims, characterized in that further reference value data for a further biometric facial feature are stored in the memory (2) which indicate a further reference value for a further geometric relationship between at least two landmarks of the facial recognition in the three-dimensional reference model for the person's face.

9. Method according to at least one of the preceding claims, characterized in that the at least one data set is stored in the memory (2) of the ID document (3) in accordance with the specifications for the logical data structure for storing biometric information in the case of machine-readable travel documents according to ICAO Doc 9303.

10. Method according to at least one of the preceding claims, characterized in that authenticity and integrity of the at least one data set are correspondingly protected by means of an electronic signature of the passive authentication for machine-readable travel documents according to ICAO Doc 9303.

11. Method according to at least one of the preceding claims, characterized in that the reference value data are determined using the model data.

12. Method according to at least one of the preceding claims, characterized in that the reference value data are determined using the landmark data.

13. Method according to at least one of the preceding claims, characterized in that the current image data for the person's face are recorded when the person presents the ID document for identification.

14. Mobile device (5) for identifying a person on the basis of a biometric facial feature, comprising: - at least one processor (6); - a data communication device (4) and - an image recording device (7); wherein the mobile device (5) is configured for the following: - reading at least one data set from the memory (2) of an ID document (3), wherein the at least one data set is stored in the memory (2) for personalizing the ID document (3) for a person to whom the ID document (3) is assigned and comprises at least one of the following data sets: - reference value data which indicate a plurality of reference values for a geometric relationship for at least two landmarks of the facial recognition in a three-dimensional reference model for the person's face, - landmark data which indicate landmarks for facial recognition in the three-dimensional reference model and from which the reference value data can be determined, and - model data which represent the three-dimensional reference model and from which the reference value data can be determined; - providing the reference value data in the at least one processor (6); - recording current image data for the person's face using the image recording device (7); - determining a current three-dimensional model for the person's face using the image data; - determining comparison data from the current three-dimensional model, wherein the comparison data indicate a plurality of comparison values for geometric relationships between the at least two landmarks of the facial recognition in the current three-dimensional model for the face, and providing the comparison data in the at least one processor (6); and - identifying the person if it is determined by means of the at least one processor (6) that at least one of the comparison values is equal to a corresponding reference value within a predefined error tolerance; characterized in that the reference values are weighted reference values and, when identifying the person, are checked in accordance with a weighted sequence one after the other, wherein the checking sequence is aborted if a correspondence with a respectively assigned comparison value within a respective tolerance range is established for one or more of the reference values.