RU2636136C2 - Portable device of biometric authentication with one-pixel sensor - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: device for biometric identification of a user comprises an illumination unit configured to emit, at least, one first optical radiation per human body; a modulation unit configured to modulate, at least, one second optical radiation scattered from a human body; a sensor configured to detect the integrated power of, at least, one second optical radiation scattered from the human body; a processing and storage unit configured to switch the states of the modulation unit to receive a signal from the sensor such that each sensor reading corresponds to a particular state of the modulation unit, performing biometric user authentication.

EFFECT: providing reliable authentication and portability of the device.

8 cl, 11 dwg

Description

FIELD OF TECHNOLOGY

The present invention relates to devices and methods for authenticating people, in particular, to an authentication technique based on the use of personal biological data, namely, configurations of veins of the human body.

In particular, the invention relates to a wearable system and method for obtaining biological information based on the configuration of veins in the wrist region of a user and performing biometric authentication of the user.

KNOWN LEVEL OF TECHNOLOGY

Currently, there are many ways to authenticate a user, which are based on the use of biological information, for example, a fingerprint, iris and retina, configurations of veins in the palm of the hand or area of the wrist. Given the high pace of distribution of wearable electronics, it is very important to provide the user with personal authentication tools for cases of using wearable devices such as watches, activity tracking devices (trackers), etc. In such cases, the wrist area is considered the only part of the human body that is suitable for obtaining biological information. This makes it possible to authenticate the user based on the configuration of the veins in the wrist. There are various solutions that rely on using vein configuration information as the basis for performing authentication procedures. However, all the described devices have some disadvantages that impede the successful and reliable use of these devices. All described devices can be divided into two groups: devices that register and process images of the vein configuration, and devices that register and process a certain signal associated with the configuration of veins. The devices of the first group work with images of the configuration of human veins, which leaves the opportunity for theft of these images. The devices of the second group are more reliable from the point of view of safety, but they are less accurate due to their high sensitivity to the displacement of these devices on the wrist.

Various inventions are currently known pertaining to the field of human authentication using biological information.

The first known solution is described in the application US 20140294251 A1. The described device registers an optical image of the configuration of human veins, extracts the characteristic features of this configuration and performs the authentication procedure. The disadvantage of this device is that it works with the image of the configuration of veins, in case of theft of which (image) the thief can gain access to the user's personal data.

A different approach is demonstrated in a device and method intended for use in a wearable form factor, which include a device for obtaining biological information, a method for obtaining biological information, and a biometric device described in US 8,144,938 B2. The device is designed to obtain biological information by projecting optical radiation onto the surface of the human body in the wrist and collecting the scattered radiation by the sensor. The authentication procedure is performed by comparing the signals received by the sensor with the signals stored in the memory. This device does not directly use vein configuration images, which is more appropriate when working with personal data compared to the device described above.

The disadvantage of this system is that it is sensitive to authentication errors that can occur if there is even a slight displacement of the device in the wrist area.

The closest features of the proposed solution can be found in the method and device for the compressed image forming apparatus claimed in US 8199244 B2, which represent a significant improvement in the field of imaging in infrared (IR) light. The imaging (imaging) device directly receives random projections of the incident light and restores the image using a single sensor. Thanks to this sensor, the device can be adapted to imaging at wavelengths to which conventional CCD and CMOS sensors are not sensitive. This is very important for biometric authentication based on recognition of vein configurations, since this task requires the use of infrared radiation due to its ability to penetrate into the human body at great depths.

The main disadvantage of this device is that it cannot be adapted to a wearable form factor due to the bulky light-modulating device and optical subsystems designed to direct the modulated light flux to photosensitive elements.

SUMMARY OF THE INVENTION

The present invention is intended to provide a compact and reliable solution to the problem of biometric authentication of a user portable enough to fit in a wearable form factor. Analysis of the known level shows that there are three main problems that the present invention solves:

- Wearable devices for user authentication use real images of vein configurations, which is not a safe practice, since these images, which are the user's personal data, can be stolen and subsequently used to gain unauthorized access to the user's device.

- Wearable devices for user authentication, which operate on signals obtained from biological data, without registering a real image of the vein configuration, are susceptible to authentication errors that may occur as a result of the device being displaced in the wrist area.

- The complexity of implementing a single-pixel camera in the form factor of a wearable device due to the cumbersome modulation unit and the optical subsystem.

The technical result of the present invention is to increase the degree of security during biometric authentication of the user of the wearable device by eliminating the need to store real images of the configurations of the veins of the user, as well as ensuring stable operation of the authentication system when the device is offset. This can be achieved by solving the three problems described above.

The first problem is solved by using a signal that is associated with biological parameters (for example, the configuration of the veins in the wrist), presented in this one-dimensional format; this excludes the possibility of reconstructing a real image of the vein configuration without a priori knowledge of the parameters of the modulation unit used to modulate the optical radiation scattered from the surface of the human body.

The second problem is solved by using a modulation unit and a processing unit that performs a light modulation procedure by switching between the masks created by the modulation unit according to a predetermined sequence selected in such a way as to compensate for the possible displacement of the device on the user's wrist.

To solve the third problem, it is proposed to use a flexible sensor, which can be made on the basis of a film solar panel attached directly to the surface of the modulation unit in such a way as to collect all the optical radiation passing through it. The modulation unit can also be implemented as a flexible film with predetermined light modulation parameters.

According to a first aspect of the present invention, there is provided an optical device for biometric user authentication, comprising:

- a lighting unit configured to emit at least one first optical radiation to the human body;

- a modulation unit configured to modulate at least one second optical radiation scattered from the human body by changing its state, characterized by an optical transfer function;

- a sensor configured to detect the integrated power of at least one second optical radiation scattered from the human body and passing through the modulation unit;

- a processing and storage unit, configured to switch the states of the modulation unit in accordance with the order selected to compensate for the possible displacement of the device; receiving a signal from the sensor so that each sensor reading corresponds to a particular state of the modulation unit; performing biometric user authentication by comparing the received signal with a reference signal associated with the user and stored in memory.

According to one embodiment of the present invention, a modulation unit configured to modulate at least one second optical radiation scattered from the human body by changing its state, characterized by an optical transfer function, is a liquid crystal display (LCD).

According to another embodiment of the present invention, a modulation unit configured to modulate at least one second optical radiation scattered from the human body by changing its state, characterized by an optical transfer function, is a diffuse diffuser with an uncontrolled optical transfer function.

According to another embodiment of the present invention, a modulation unit configured to modulate at least one second optical radiation scattered from the human body by changing its state, characterized by an optical transfer function, is a diffuse diffuser with a controlled optical transfer function.

According to another embodiment of the present invention, the sensor is a single point sensor.

According to another embodiment of the present invention, the sensor is a photovoltaic cell.

According to another embodiment of the present invention, the sensor is a photodiode.

In accordance with another aspect of the present invention, a method for biometric authentication of a user using a device for biometric authentication, comprising the steps of:

- at least one first optical radiation is emitted to the human body;

- modulation of at least one second optical radiation scattered from the human body, by passing it through a modulation unit, which changes its state, characterized by an optical transfer function, in a predetermined order, established and controlled by the processing and storage unit, and a predetermined order selected to compensate for possible displacement of the device;

- the integrated (total) power of at least one second optical radiation scattered from the human body and passing through the modulation unit is determined by the sensor;

- the signal from the sensor is recorded so that each reading from the sensor corresponds to a specific state of the modulation block;

- biometric authentication of the user is performed by comparing the received signal with a reference signal associated with the user and stored in memory.

According to another embodiment of the present invention, the modulation of at least one second optical radiation scattered from the human body is carried out by passing said radiation through a modulation unit with an uncontrolled optical transfer function.

According to another embodiment of the present invention, the modulation of at least one second optical radiation scattered from the human body is carried out by passing said radiation through a modulation unit with a controlled optical transfer function.

According to another embodiment of the present invention, authentication is considered positive when the correlation coefficient between the real signal and the reference signal exceeds a predetermined level.

According to another embodiment of the present invention, a reference signal associated with a user is previously stored in memory upon initial calibration of the device.

The main differences from the prototype include the use of a modulation unit in combination with a compact and flexible optical radiation sensor, and a processing unit that performs the switching procedure between the masks formed by the modulation unit according to a predetermined order selected in such a way as to compensate for the possible displacement of the device on the wrist user. These differences make it possible to obtain biological information and perform stable user authentication with wearable devices without the need to save real images of vein configurations in the device’s memory.

In an exemplary embodiment of the invention, the surface of the wrist region is illuminated with an incoherent light source; incident light is scattered from the tissue of the human body and propagates through the modulation unit. During the authentication procedure, the modulation unit generates a binary mask in accordance with the signals of the control unit and modulates the spatial distribution of the propagating light. At the output of the modulation unit, all the power of the outgoing light is collected and recorded by a single-point sensor. The time dependence of the sensor signal is stored in the device memory for the authentication procedure. The authentication procedure is performed by comparing the described dependencies stored in the device memory and received at the time of switching the device. With a sufficiently high correlation coefficient (> 0.95) of these dependencies, authentication is recognized as positive.

BRIEF DESCRIPTION OF THE DRAWINGS

Other details and advantages of the invention will be described below with reference to the accompanying drawings.

FIG. 1 depicts a general diagram of a wearable device for biometric user authentication, which shows the blocks contained in the invention and their functional relationships;

FIG. 2 is a diagram of a preferred embodiment of a wearable device for biometric user authentication;

FIG. 3 shows a typical binary mask formed by a modulation block;

FIG. 4 shows the dependence of the correlation coefficient between the sensor signals recorded in the case of a different offset value of the wearable device in the side plane;

FIG. 5 represents images of the configuration of veins in the wrist region in the event of a displacement of the wearable device;

FIG. 6 depicts binary masks shifted relative to each other;

FIG. 7 shows the signals recorded by the sensor in the event of a displacement of the wearable device;

FIG. 8 is an image of the configuration of veins in the wrist;

FIG. 9 is an enlarged view of the configuration of the veins in the wrist;

FIG. 10 is an enlarged image of a wrist configuration of veins reconstructed using a sensor;

FIG. 11 depicts a flowchart of a biometric authentication method for a user.

DETAILED DESCRIPTION OF EMBODIMENTS OF THE INVENTION

A block diagram of a standalone optical device for biometric authentication shown in FIG. 1, displays the interaction between the various units of the device, where 1 is the lighting unit, 2 is the first optical radiation, 3 is the study area, 4 is the second optical radiation, 5 is the modulation unit, 6 is the sensor, 7 is the processing and storage unit. The device shown as an example includes a lighting unit 1, which may consist of one or more sources of the first optical radiation 2 with different wavelengths illuminating the study area 3 (for example, the wrist region of the human body). The first optical radiation penetrates the tissue of the human body, is reflected and scattered in the opposite direction. Optical radiation propagating in the opposite direction from the human body in the studied region 3 is hereinafter referred to as the second optical radiation 4. The second optical radiation 4, which may have components with different wavelengths depending on the parameters of the lighting unit 1 (different sources of optical radiation with different lengths waves for greater reliability), is distributed through modulation unit 5, which forms special masks and switches between them to perform modulation of the spatial distribution of the intensity of the propagating light according to the control signals from the block 7 processing and storage. Special masks of the modulation unit 5 represent various spatial distributions of the optical parameters of the medium through which the second optical radiation 4 propagates. Such masks modulate the second optical radiation 4 by changing its optical parameters (for example, the spatial distribution of amplitude, phase, etc.). At the output of the modulation unit 5, the total power of the modulated light is detected by the sensor 6. As a result, the detected biological signal represents the time dependence of the total power of the detected light, as shown in FIG. 4. This signal is stored in the memory of the processing and storage unit 7 and is used to perform the authentication procedure, which involves matching the registered biological signal and the signal stored in the memory after the initial calibration, which will be described below. If the correlation coefficient between these signals is high enough (for example,> 0.95), authentication is considered positive. To compensate for the possible displacement of the device, a special procedure is used, described below.

According to the preferred embodiment shown in FIG. 2, all components of the system are placed in a wearable device that the user puts on his wrist, where 8 is a light source, 9 is a cylindrical lens, 10 is an external illumination device. When the device is put on, and the user wants to initialize it and gain access to its basic functions, he must perform the authentication procedure. This procedure begins with a user pressing a special button, touching the touch screen, etc. The lighting unit 1, containing a light source 8, a cylindrical lens 9 and an external illumination device 10, starts to emit the first optical radiation, which is incoherent, and whose wavelength is selected so as to ensure maximum penetration into body tissues. The lighting unit 1 may comprise light sources with wavelengths in the near infrared range of 750-950 nm. This radiation is collected by a cylindrical lens 9 and focuses on the side surface of the external illumination device 10, which is compact enough to fit in the form factor of the wearable device. The external illumination device 10 creates close to uniform illumination of the surface of the human body in the wrist. The first optical radiation penetrates the human skin in the wrist and is scattered mainly on red blood cells in the blood vessels and thereby forms the reflected second optical radiation with a high enough intensity so that contrast images of the configuration of veins located in the wrist can be obtained. The second optical radiation passes through a modulation unit 5, which is, for example, a liquid crystal display (LCD) controlled by a sequence of signals from the processing and storage unit 7. Each signal from the received sequence switches the LCD to a state in which the spatial amplitude function of the light transmission represents a two-dimensional binary mask similar to that shown in FIG. 3, where black pixels correspond to zero light transmission, and white pixels to full light transmission. The total output power of the light passing through the LCD with the binary function of transmitting light is detected by the sensor 6 so that for each binary mask of the LCD there is a single value of the sensor signal. Thus, the complete signal detected by the sensor 6 during one recording cycle is similar to the reference signal shown in FIG. 4 (this figure shows the dependence of the total output power detected by the sensor on the binary mask number generated by the modulation unit). This procedure is performed every time authentication is required. Authentication, as such, is performed by comparing the current complete signal from the sensor 6 with the signal stored in the device memory. As a measure of the correspondence of these signals, a correlation coefficient can be used. When the correlation coefficient exceeds a predetermined level, for example 0.95, authentication is considered positive. This approach completely eliminates the storage in the memory of real images of the configuration of the veins of the user and makes it impossible to steal this data and unauthorized access to the wearable device.

The proposed method also allows to solve the problem of possible displacement of the wearable device on the wrist. Suppose the device is worn on the wrist not exactly in the same place as before. In this case, the vein configuration structure shifts as shown in FIG. 5. The displacement of the vein configuration changes the signal detected by the sensor 6. With a small value of the displacement, the reference signal (Fig. 4) can turn into the signal shown by the dashed line in Fig. 4. This figure shows that these signals are still highly correlated (the correlation coefficient exceeds 0.95), therefore, authentication will be positive. However, the offset value of the device may be higher so that these signals are not correlated. To overcome this problem, the described method is proposed. Each binary mask from the full sequence is displayed on the LCD so that the spatial distribution of its black and white pixels is slightly biased, as shown in FIG. 6. This procedure is repeated several times by “shifting” each binary mask from the complete sequence in various directions to a different value. Thus, we obtain the dependence of the correlation coefficient between the reference signal and the signals measured for various shifts of binary masks along the X and Y directions in the LCD plane. This dependence can be represented as a function of two variables — shifts along the X and Y axes, or a matrix, and can be visualized in the form of a two-dimensional graph shown in FIG. 7. As can be seen from the graph, the maximum value of the correlation coefficient is achieved at a point with coordinates (-5, -8) relative to the center of the plot with coordinates (0,0). The length of the black arrow is proportional to the value of the shift, and its direction corresponds to the direction of the shift. Therefore, in order to compensate for the possible bias during the authentication procedure, it is necessary to obtain a matrix consisting of the values of the correlation coefficients between the reference and biased signals and check for the presence of values exceeding the specified value (for example, 0.95). If such a value is found, then the authentication procedure is considered positive.

The second variant of the device involves the use of modulation unit 5, which is a diffuse diffuser with a controlled optical transfer function. As a diffuse diffuser, you can use a thin plastic film with optical parameters that are determined during calibration at the device manufacturing stage. In this case, the masks formed by the modulation unit 5 are manually changed, for example, by means of a spring suspension of a thin film — the modulation unit 5, without receiving signals from the processing and storage unit 7. This scheme provides a higher degree of security, since the modulation unit 5 can be manufactured with randomly selected optical transfer functions that are unique to each device. However, this solution is more vulnerable to bias errors; these errors must be overcome by using a modulation unit 5 with an optical transfer function with a periodic spatial distribution. Another way to compensate for bias errors is to reconstruct a real image of the vein configuration in the wrist area to obtain numerical data about the direction and magnitude of the bias of the device. This data is necessary for the correct placement of the device followed by the authentication procedure. The image reconstruction problem is solved using the sequence b (t) measured by the sensor and represents the solution to the “base pursuit” problem:

, (1)

, (one)

- L1-норма вектора х, A - матрица известных коэффициентов, представленных бинарными масками, сформированными блоком модуляции. Типичное изображение конфигурации вен, которое подлежит реконструкции, представлено на фиг. 8 и 9 (увеличенное изображение конфигурации вен). Результат реконструкции представлен на фиг. 10. Можно заметить, что реконструированное изображение отображает все признаки, представленные в исходном изображении, и может быть использовано для распознавания конфигурации вен.where x is a vector representing the spatial distribution of light in the plane of the modulation block,

- L1-norm of the vector x, A is a matrix of known coefficients represented by binary masks formed by the modulation block. A typical vein configuration to be reconstructed is shown in FIG. 8 and 9 (enlarged image of the configuration of the veins). The reconstruction result is shown in FIG. 10. You may notice that the reconstructed image displays all the features presented in the original image and can be used to recognize the configuration of veins.

To increase the security level of the device, you can use the modulation unit 5, which is a diffuse diffuser with an uncontrolled optical transfer function. As a diffuse diffuser, you can use a thin layer of a translucent liquid (for example, any liquid with solid dispersed particles) in a flat transparent container. Dispersed particles serve as separate light scatterers and provide random transmission of light through the medium. Therefore, the masks generated by the modulation unit 5 cannot be repeated. In this case, to extract data on the configuration of the veins, the condition must be met - the total power of the signal passing through the modulation unit 5 must remain unchanged.

The algorithm shown in FIG. 11 shows the essence of the proposed method for continuous measurement of at least one physiological parameter of a person. At least one first optical radiation is directed to the human body in the study area (for example, the wrist area). It scatters, forming a second optical radiation, the spatial distribution of which carries data on the configuration of the veins in this illuminated area. The second optical radiation passes through a modulation unit, which modulates its spatial distribution, by generating a set of masks that are switched on in series. The total (integral) light power at the output of the modulation unit is recorded by the sensor in such a way that each mask created by the modulation unit is responsible for one sample of the complete signal recorded by the sensor. The sensor signal is recorded and processed by the processing and storage unit, which performs the biometric authentication procedure - compares the real signal from the sensor with one reference signal stored in the memory of the processing and storage unit; in the case of a high predetermined degree of coincidence of the signals, the authentication procedure is considered positive.

A user scenario according to a preferred embodiment can be described as follows. Before the first use of biometric authentication using a wearable device, the user must perform the initial calibration of the device - save personal biometric data in the device’s memory. To do this, the following procedure is performed. The user puts the device on his hand in the wrist. After its inclusion, the lighting unit begins to emit the first optical radiation, which illuminates the human skin and is scattered, thereby creating a second optical radiation. It is distributed through a modulation unit, which immediately begins to form binary masks in accordance with the control sequence from the processing and storage unit. The formation of a complete set of binary masks in a modulation unit is expected to take about 1 second. Modulated optical radiation after it passes through the modulation unit is detected by the sensor. The signal from the sensor output is recognized as a reference and stored in the memory of the processing and storage unit. This completes the initial calibration.

For the first biometric authentication, the above procedure is repeated, but in this case, the received signal is compared with the reference signal stored in the memory of the processing and storage unit. If a high degree of coincidence between the reference and registered signals is achieved (for example, the correlation coefficient between these signals is higher than 0.95), then the authentication procedure is completed and is considered positive. This means that the owner of this wearable device is verified, and on this basis all the capabilities of the wearable device are unlocked. After successful authentication, the user can receive audio or video through the built-in audio device or through the device’s display. Otherwise, the device will remain locked.

Claims (18)

1. An optical device for biometric authentication of a user, comprising: - an irradiation unit configured to emit at least one first optical radiation to the configuration of the veins of the human body in the wrist; - a modulation unit configured to modulate at least one second optical radiation scattered from the human body by changing its state characterized by an optical transfer function by forming a set of two-dimensional binary masks and sequentially switching them to modulate the spatial intensity distribution of the propagating second optical radiation; - a sensor configured to detect the integrated power of at least one second optical radiation scattered from the human body and passing through the modulation unit; - a processing and storage unit, configured to perform biometric authentication of the user by comparing the received signal with a reference signal associated with the user and stored in memory, to form a matrix consisting of the values of the correlation coefficients between the reference and biased signals measured for various shifts of binary masks along the X and Y directions in the plane of the modulation block, and detect the presence of correlation coefficients in excess of a given value. 2. The optical device according to claim 1, in which a modulation unit configured to modulate at least one second optical radiation scattered from the human body by changing its state, characterized by an optical transfer function, is a liquid crystal display (LCD). 3. The optical device according to claim 1 or 2, in which the sensor is a single point sensor. 4. The optical device according to claim 1 or 2, in which the sensor is a photoelectric element. 5. The optical device according to claim 1 or 2, in which the sensor is a photodiode. 6. A method for biometric authentication of a user using a device for biometric authentication, comprising the steps of: - emit at least one first optical radiation on the configuration of the veins of the human body in the wrist; - modulate at least one second optical radiation scattered from the human body by passing it through a modulation unit that changes its state, characterized by an optical transfer function, by forming a set of two-dimensional binary masks and sequentially switching them to modulate the spatial distribution of the intensity of the propagating the second optical radiation in a predetermined order established and controlled by the processing unit and storage, and a predetermined order is selected so as to compensate for the possible displacement of the device; - determine the sensor integrated power of at least one second optical radiation scattered from the human body and passing through the modulation unit; - receive a signal from the sensor so that each reading from the sensor corresponds to a specific state of the modulation unit; - for biometric authentication of the user by comparing the received signal with a reference signal associated with the user and stored in memory, a matrix is formed consisting of the values of the correlation coefficients between the reference and biased signals measured for various shifts of binary masks along the X and Y directions in the block plane modulation, and detect the presence of correlation coefficients in excess of a given value. 7. The method according to claim 6, in which authentication is recognized as positive when the correlation coefficient between the received signal and the reference signal exceeds a predetermined level. 8. The method according to claim 6, in which the reference signal associated with the user is previously stored in memory during the initial calibration of the device.

Images