“System for secure acquisition of photographs of human faces”

DESCRIPTION

Field of the invention

The present invention relates to a system for secure acquisition of photographs of human faces.

Background art

In order to issue an identity document, a license or a particular permission, competent authorities need an official photograph of the applicant’s face. Such photograph must meet some acceptance requirements, i.e. it must comply with very precise and detailed international specifications, such as the ICAO 09303 international standard.

Because of a particular criminal phenomenon called “MORPHING”, which will be briefly described below, institutions such as Embassies, General Register Offices, Departments of Motor Vehicles, Police Headquarters and other authorized bodies are currently tending to no longer accept common pre-printed “passport photos” that applicants normally hand in when requesting a document.

The need is therefore felt for providing every document-issuing office with a specific technology that will allow for “live” acquisition of the applicant’s photo and that will ensure full compliance and compatibility with the laws currently in force.

The difficulty in acquiring a standard-compliant photograph is related to a number of technical/practical aspects. The technical/qualitative characteristics that an ICAO-compliant document photograph must have are summarized below:

• the face must cover 70-80% of the photograph, from the chin base to the forehead;

• image quality (resolution);

• even illumination;

• absence of shadows on the face;

• area of the photographed face;

• face position;

• color photo;

• neutral expression and closed mouth; • open, clearly visible eyes;

• white background with uniform light; no flash reflections on the face and elimination of the “red-eye” effect;

• focusing suitable for obtaining the required sharpness;

• the person must stare at the IDBOX;

• size: 35-40 mm;

• printing on high-quality, high-definition paper;

• well-centered face, not in profile;

• photo without any other objects or people in the background.

“MORPHING” is one of the first digital effects developed by the motion picture industry, and consists of a fluid, gradual and continuous combination of two different images, which may be objects or, as is the case of the present invention, faces.

The fraudulent “Face Morphing attack” uses “morphed” images obtained by combining faces belonging to different identities. This attack envisages that, when a person’s identity is checked, the image saved in the document will correspond to the live-acquired one of more than just one person, so that not only the actual holder, but also other people can be identified using the same document (up to 8 different identities). The basic idea of this attack is that a person, who is being sought by the police or is not permitted to leave the country of origin, finds a person with a face resembling his/hers.

The issued document is perfectly regular, but nonetheless fraudulent. In technical jargon, such counterfeited documents are called “FODs”, Fraudulently Obtained Documents, or “FOGs”, Fraudulently Obtained Genuine documents.

The fraudulent action does not lie in an alteration of the contents of the document, but in an attempt to deceive the officer who is in charge of issuing the document by giving him/her a pre-printed morphed photograph, thus acting in a manner that is equivalent to a normal application for a document (passport, identity card, driver’s license, visa/residence permit, etc.).

The issued document will pass all integrity checks at any control gate.

It has been demonstrated that:

1 It is possible to create a realistic image by morphing;

2 The transformed image can deceive the inspector;

3 The most advanced face recognition algorithms can be deceived. A variety of systems and methods are currently known in the art for the photographic acquisition of a person’s face, as required for obtaining any type of identity document. One of the most common ones is the use of paid services provided by a professional photographer, who offers his/her subjective competence to acquire the best photo of the requesting person. The photographer must observe the above-described ICAO specifications but, since during the procedure carried out on the photographer’s premises no live monitoring system is available and the final user of the photograph is not compelled to verify the compliance of the photograph with the ICAO standard, the compliance of the photograph with the standard can only be inferred by the photographer or the final user by visual deduction.

This relationship between customer and service provider is essentially based on mutual trust. While it is known that the photograph of a face is a very important personal biometric datum, which is attached to identity documents anywhere in the world, and that nowadays allows crossing customs barriers automatically (e-gates) or under manual verification by an operator, accessing banking, insurance, digital identity, etc. services, or that, in general, can be used by police forces to identify a person for civil or criminalistic purposes. If such datum is not supported by certified data and consistent procedures at national and European level, some room will certainly be left for fraudulent behaviours by ill-intentioned people (e.g. morphing).

Several automatic systems for face image acquisition are also known, such as photo booths spread over the territory. In this case as well, however, no procedures are explicitly applied for verifying the compliance of the photo with the law in force.

Lastly, it should also be taken into account that it is very easy for anyone to take a passport photo necessary for obtaining an identity document by themselves, using common and accessible technological systems available on the market, such as PCs, smartphones, tablets, which are now equipped with high-performance cameras and software capable of simulating the skills of a professional photographer. This latter mode is certainly most suitable, also due to the huge number of editing applications available, for manipulating the photographic datum in order to deliberately disguise one’s identity.

Summary of the invention

It is therefore the object of the present invention to provide a system for secure acquisition of photographs of human faces which can overcome all of the above problems. The system for secure acquisition of photographs of human faces according to the invention aims at eliminating the human error or manipulation component by fully transferring skills and requirements to technology.

The system of the invention generates a high-quality digital image of a human face which is compliant with the following international directives and guidelines:

• Annex - Commission Implementing Decision C(2018) 7774 of 30 November 2018;

• Annex 3 - Commission Implementing Decision C(2018) 7767 of 30 November 2018;

• ICAO Technical Report Portrait Quality (reference facial images for MRTD) v.1 ;

• ICAO 9303 7th Edition;

• ISO 19794-5:2011 - Frontal Image Type;

• Commission Implementing Decision (EU) 2019/329 of 25 February 2019.

The system is ergonomical and easy to use, and can be adapted to public administration offices, minimizing any impact on logistics and areas; it represents a stable and robust solution, and is preferably provided with an anchorage system to ensure the utmost safety.

The present invention relates to a system for secure acquisition of photographs of human faces which comprises:

- two or more digital cameras, adapted to acquire the image of a face, said cameras being of the color type and having an angular aperture and a sensitivity such as to obtain a minimum number of pixels within the interocular distance of the face;

- an illumination system comprising LED illuminators;

- an aiding system for correctly positioning the face, comprising luminous indicators arranged around said two or more digital cameras and at least one loudspeaker;

- an electronic control, power supply and hub circuit of said system;

- one or more first column-type rigid supports incorporating said digital cameras, and adapted to keep said two or more cameras in a vertical mutual position at a given mutual distance;

- one or more second column-type rigid supports for said illumination system, said second column-type rigid supports being connected in front of said one or more first column-type rigid supports in such a way that they can rotate between a position in which they hide said cameras and a position in which said illuminators are located at the sides of said cameras;

- at least one propping and containment support for said system, adapted to support said system in the vertical position;

- said electronic circuit comprising the following software means, adapted to operate in succession:

- means for selecting one camera of said two or more digital cameras which is most suitable for acquiring the photograph of the person’s face, the selected camera being the one with respect to which the face is centered at best;

- means for activating said aiding system for correctly positioning the face in front of said selected camera, providing visual indications by activating said luminous indicators and audible indications emitted through said at least one loudspeaker, adapted to have the face move and center itself in front of said selected camera, at the correct distance from said selected camera;

- means for measuring the brightness of the environment in which said face is located, and for activating said illumination system to obtain the utmost uniformity in the illumination of said face;

- means for taking a photograph of said face when said means for activating said aiding system provide an indication that the face is centered correctly.

It is a particular object of the present invention to provide a system for secure acquisition of photographs of human faces as described in detail in the claims, which are an integral part of the present description.

Brief description of the drawings

Further objects and advantages of the present invention will become apparent from the following detailed description of a preferred embodiment (and variants) thereof, wherein reference will be made to the annexed drawings, which are only supplied by way of nonlimiting example, wherein:

Figures 1.1 to 4 show some examples of embodiment of the system according to the invention;

Figures 5 and 6 show some possible arrangements of the system towards persons whose faces have to be photographed;

Figure 7 shows one possible embodiment of the aiding system for correctly positioning the face around a camera;

Figure 8 shows one possible arrangement of the system on a desk. In the drawings, the same reference numerals and letters identify the same items or components.

Detailed description of some embodiments

The following will describe some examples of embodiment of the system for secure acquisition of photographs of human faces according to the invention. Such embodiments are only illustrated herein by way of example, since many variations thereof are still possible. With particular reference to Figures 1.1 - 4, the collaborative photograph acquisition system for taking a photograph of a person’s face for enrollment purposes basically comprises:

• two or more digital cameras 1, which acquire the face image;

• an LED illumination system 2;

• an aiding system for correctly positioning the face 3;

• an electronic control, power supply and hub circuit;

• column-type rigid supports 5 and 6 for, respectively, the digital cameras 1 and the illumination system 2;

• a propping and containment support 4.

The various components are described below.

Two preferable implementation solutions are shown:

A first solution comprises two digital cameras 1 in column (Figures 1.1, 2.1), preferably for desk application, and a second solution comprises three or four digital cameras 1 in column (Figures 1.2, 3, 4), preferably for floor application. The cameras are integrated into one or more column-type rigid supports 5, while the LED illumination system 2 comprises pairs of emitters positioned at the ends of further column-type rigid supports 6. The cameras are disposed in a mutual vertical position, so as to facilitate the use of the system in front of people of different height, who can thus, advantageously, stay in the same position.

Preferably, the column-type rigid supports 5 and 6 are interconnected at the center, so that the rigid element 6, which supports the LED illumination system 2, can rotate in front of the rigid element 5, which supports the cameras 1, between two positions, i.e. an open position with an angle of less than 90° (Figures 3, 4), which reveals the cameras 1, and a front position (Figure 2.1), which hides the cameras 1. The angle of the open position is such that the illuminators are positioned at the sides of the cameras to obtain a correct parallax and avoid a negative frontal illumination of the face. Multiple pairs, e.g. two pairs (Figures 1.2, 3, 4) of rigid elements 5 and 6 can be connected in column, so as to make it possible to take photographs correctly of faces of people of different height.

Two types of system support 4 are preferably provided: one (4’) for fastening it to a desk (Figures 1.1, 2.1, 8), and another (4”) with a floor-stand for supporting the system on a floor (Figures 1.2, 2.2, 3, 4).

With reference to Figure 7, the aiding system for correctly positioning the face 3 comprises luminous visual indicators, preferably eight of them, arranged around each camera 1.

Preferably, a loudspeaker 8 is also included, which is incorporated into a column-type rigid support 5. In one variant, it may be incorporated into a column-type rigid support 6.

The luminous visual indicators and the loudspeaker are adapted to guide the person in real time to assume the correct face and eye angle in order to ensure an optimal photographic result.

The LED illumination system 2 improves the photographic acquisition by illuminating the face of the person standing in front of the system.

The electronic control, power supply and hub circuit is integrated into the system, e.g. into the rigid element 5, and comprises software that makes it possible to control:

• the light (LED illumination system);

• the cameras;

• the aiding system for correctly positioning the face;

• the ICAO verification of the photo.

The system is modular; therefore, it is a single object that can take different configurations. As aforesaid, it is possible to mutually join two systems by stacking and electrically connecting them to obtain a device with three or more cameras for acquiring a photograph of a standing person’s face. The basic principle of operation is to have a column/totem of cameras for covering the required shooting height and inclination ranges. While the person is standing in front of the system, the arrows of the aiding system for correctly positioning the face 3 will light up in green or red depending on whether the face is positioned correctly or not, for the purpose of guiding the person to assume the correct face position. The arrows will become green when the criteria required by the ICAO standard for document photos are met, and at that point the person will have to keep his/her face in that position. A photograph will then be taken, which will then be automatically transferred to a storage system, e.g. held by the authority. Along with the visual indications, a voice synthesizer included in the electronic circuit will reproduce the same indications through the loudspeaker 8, in different languages, to facilitate the understanding of the emitted visual signals and also as an aid for blind people.

Figure 4 shows three views of the system in the variant with a floor-stand and two pairs of stacked column-type rigid elements 5 and 6.

Figures 5 and 6 show two practical applications of the system in front of persons whose face has to be photographed, respectively in a standing position with the floor-stand variant of the system and in a seated position with the desk variant of the system.

In order to generate a photograph compliant with the ICAO standard, the software can:

• integrate itself with the aiding system for correctly positioning the face, in order to suggest the correct face and eye position to the person;

• integrate itself with the LED illumination system in order to control the luminous intensity according to the brightness of the environment and the color of the person’s skin;

• integrate itself with the cameras.

Merely by way of non-limiting example, some indications about the specifications of the various elements that make up the system are given below.

Weight and dimensions of the individual components

Total weight: 1.6 Kg

• Cameras/illuminators module: 800gr - 54.5cm x 8.5cm x 7.54 cm (height x width x depth)

• Desk tripod: 800gr, maximum size when closed 49.5cm, when open, it occupies an area of 40x40cm

• Bracket with desk clamp: 800gr, height 49.6cm

Materials used: ABS and Aluminum.

As concerns the electronic control, power supply and hub circuit, it controls the light panels and the face positioning aiding panels, and is structured to perform the following functions:

• Generation of the voltages necessary for the operation of the system, so that the power available at the input USB port can be used for enabling the power supply to the whole system, which is derived from the 24VDC input.

• Delivery of the 5VDC power supply to the 1mA USB input. • Management and control of the light modules, i.e. of the luminous intensity of the light modules, by pulse width modulation (PWM).

• USB hub for all USB ports - Cameras, Host USB port, control board USB port, voice module that permits sending voice messages to the user to allow him/her to properly position his/her face, are all connected to this hub.

• Management of the device control protocol.

• It hosts and controls the voice module.

As concerns the LED illumination system 2, the following values may be used:

As concerns the digital cameras, they are preferably of the 12MP or higher AF color camera type, with FOV of 75°, with a mutual distance of approximately 35 cm. More generally, different values of such quantities are possible to meet the requirements of the ICAO standard, which envisage that the sensitivity of the cameras must be such as to ensure the generation of a minimum number of pixels within the interocular distance of the face of the person to be photographed, also in accordance with the distance of the person from the acquisition system, which is preferably in the range of 70 cm to 100 cm.

As concerns the luminous visual indicators 2 positioned around the cameras, they may be panel-shaped LED indicators with variable green/red light.

The system comprises software means capable of activating the following procedure when interacting with a user:

1) The system selects a camera which is most suitable for acquiring the photograph of the person’s face depending on the person’s height (e.g. 100 to 200 cm). There are no moving parts and there is no need for an operator to position the camera at the correct height, since the software executes a procedure for checking the presence of a face in the video frame of the various cameras of the device and then activates that camera which is most suited to ensuring an ICAO-compliant acquisition, i.e. the one with respect to which the face is centered at best.

2) The system provides the person with visual and voice indications for properly positioning his/her face in front of the camera. By software processing, the integrated luminous indicators and loudspeakers guide the person in taking the correct position. For example, the red color indicates an incorrect position that needs to be changed, whereas the green color indicates a correct position that can be maintained. For example, the voice synthesizer emits, through the loudspeaker, voice indications that guide the person as he/she moves his/her head to ensure proper centering.

The two above-described procedures make it possible to overcome the following problems: a. Subjective determination, by the individual or by the operator, of the correct position of the face, which must fulfil very stringent parameters to comply with the ICAO specifications. b. Possibility for deaf and blind people to use the system autonomously. c. Possibility of giving multilingual indications to people of different ethnic groups. In fact, the voice synthesis software integrated into the system allows an operator to select different available languages.

3) Based on face position and environmental brightness, the software adjusts the intensity of the light emitted by the illumination system to ensure the best photographic acquisition. Preferably, the intensity of the emitted light is higher on the side of the illuminator which is closest to the activated camera, while one or more other illuminators that are farther from the activated camera may emit light at lower intensity, so as to avoid any shadows and obtain the utmost uniformity in the illumination of the face.

4) The software automatically verifies the correct positioning of the individual in front of the camera and the compliance of the live image of the face with all requirements of the ICAO standard, then automatically acquires the photo and stores it into the file system.

The following will illustrate some further details of the implemented software procedures. For the analysis of the image and the verification of the parameters dictated by the ICAO standard, the software uses per se known libraries, such as OpenCv (https and Dlib (http://dlib.ne ). Applying the elements contained in said libraries, the software carries out the following in succession:

- the evaluation indicated above at 1), i.e. it activates the cameras in succession and chooses the one that shows the person’s face in the most centered position relative to the angular shooting aperture;

- the evaluation indicated above at 2), i.e. it analyzes the pixels of the image received from the selected camera and evaluates the position, centering and orientation of the face while activating the visual and voice indicators as previously described herein, so as to obtain any necessary movements of the person’s face, which must be as centered and parallel to the system as possible. It also evaluates the interocular distance as a number of pixels (as described above), giving indications about the correct distance of the face from the camera. In particular, an excessively small number of pixels indicates that the person is too distant from the camera, while an excessively high number of pixels indicates that the person is too close.

- the evaluation indicated above at 3), i.e. it evaluates the illumination of the face and controls the illumination system in such a way as to obtain a proper illumination and avoid any shadows or excessive light.

- finally, it takes a photograph of the person’s face.

As previously described herein, the system is preferably available in two versions, i.e. desk type and floor-stand type.

The functionalities obtained with the desk version are essentially the following:

• optical and acoustic interaction to suggest the correct face and eye positioning to the person;

• fast acquisition operation (max. 5 seconds from the start of the procedure);

• even background behind the framed subject, in accordance with the above-mentioned standards;

• Ethernet and USB connectivity, which allows it to be physically placed anywhere;

• frontal acquisition of face images at different heights ranging from 30 cm to 110 cm;

• automatisms in the face acquisition process;

• the system can be transported in a trolley bag.

The functionalities obtained with the floor-stand version are essentially the following:

• optical and acoustic interaction to suggest the correct face and eye positioning to the person; • fast acquisition operation (max. 5 seconds from the start of the procedure);

• even background behind the framed subject, in accordance with the above-mentioned standards;

• Ethernet and USB connectivity, which allows it to be physically placed anywhere;

• frontal acquisition of face images at different heights ranging from 100 cm to 220 cm;

• automatisms in the face acquisition process.

The above-described example of embodiment may be subject to variations without departing from the protection scope of the present invention, including all equivalent designs known to a man skilled in the art.

The elements and features shown in the various preferred embodiments may be combined together without however departing from the protection scope of the present invention.

The advantages deriving from the application of the present invention are apparent, like the following.

Unlike a common webcam or digital camera, the system of the invention has, in particular, the following peculiarities:

• ease of use through a simple application;

• acquisition speed (approx. 2 seconds from the moment when the person has correctly positioned him/herself);

• automatic verification of ICAO compatibility;

• usable also by unskilled personnel;

• possibility of taking photographs of people of height ranging from 100 cm to 200 cm;

• acquisition of hundreds of photos in a few seconds, so that the best one can be chosen;

• autofocus;

• automatic face coverage control;

• color photos;

• Full High Resolution;

• inter eye distance: 240, 120 and 90 pixels;

• direct and immediate quality control to obtain ICAO-compliant photos only;

• indications about the correct positioning of the person to be photographed provided through a voice synthesizer (guiding voice in multiple languages);

• optical interaction (luminous LED indicators) to indicate the correct face and eye positioning to the person; • since it is equipped with several integrated cameras, the system automatically identifies the person’s face and enables that camera which is positioned at best;

• automatic background removal; possibility of adding a predefined background;

• the illumination system is automatic and adaptable: it automatically adjusts the amount of light according to the color of the skin, the color of the clothes, and environmental light;

• the system signals when the photo is not ICAO-compliant;

• easily transportable;

• easily connectable to any computer by means of a USB cable;

• very low consumption (it operates at a very low voltage with very low power consumption, max. 30W);

• possibility of integration into any software platform.

From the above description, those skilled in the art will be able to produce the object of the invention without introducing any further construction details.