Biometric identification device with ultrasonic transducer and proof of life detection

Abstract

Biometric identification device with ultrasonic transducer and detection of proof of life The present description relates to a biometric identification and living detection device (100) comprising at least one ultrasonic transducer circuit (104), a control circuit (106) and a data processing circuit (108), configured to implement at least the following steps: - acquisition (216) of successive depth images (164) of at least one element (101) intended to be placed against a capture surface (102) of the device (100); - detection (218), by comparison of the depth images (164) with each other, of at least one modification representative of blood circulation in a vasculature of the element (101); - determination (220) of the living character or not of the element (101) as a function of the detection or not of the modification representative of the blood circulation. Figure for the abstract: Fig. 3

Description

Title of the invention: Biometric identification device with ultrasonic transducer and proof of life detection Technical field

[0001] The present disclosure relates to the field of biometric identification devices based on capture and detection by ultrasonic transduction. Prior art

[0002] A biometric identification device makes it possible to verify or determine the identity of a user of the device based on a measurement of at least one biometric characteristic of the user such as a fingerprint, the shape of the face, the pattern of the iris, the pattern of the retina, etc. The different techniques for carrying out measurements of one or other of these characteristics each have advantages and disadvantages according to different criteria, including in particular: more or less significant error rate, ease or not of theft of the template with which the measurements are compared, possibility of detecting or not an imitation, proof or not of life, ease and comfort of use, size of the sensor required, energy consumption required, etc.

[0003] It is known to produce a biometric identification, anti-spoofing and living detection device using a fingerprint sensor carrying out optical or capacitive measurements in order to verify whether the captured fingerprint corresponds to the one expected (identity verification phase), then to implement a living detection and anti-spoofing algorithm via image processing carried out from the captured fingerprint or by using an infrared system making it possible, for example, to detect whether the finger placed on the device during fingerprint capture is a false finger (for example produced by using the image of a fingerprint present on an object touched by the person whose identity is that expected by the device).

[0004] Document US10262188B2 proposes to carry out detection of a false finger by an ultrasonic fingerprint sensor based on the quantity of energy reflected by the finger present on the sensor. This solution, however, has the disadvantage of not detecting certain fraud situations, such as for example the use of a severed finger of the user whose identity is that expected by the sensor, or even when this user is deceased. Summary of the invention

[0005] There is a need to propose a biometric identification device which does not have at least some of the drawbacks of existing solutions.

[0006] One embodiment overcomes all or part of the drawbacks of known solutions and proposes a biometric identification and living organism detection device comprising at least one ultrasonic transducer circuit, a control circuit and a data processing circuit, configured to implement at least the following steps:

[0007] - acquisition of successive depth images of at least one element intended to be disposed against a capture surface of the device;

[0008] - detection, by comparing the depth images with each other, of at least one representative modification of blood circulation in a vasculature of the element;

[0009] - determination of the living or non-living character of the element based on the detection or not of the representative modification of blood circulation.

[0010] According to a particular embodiment, the ultrasonic transducer circuit, the control circuit and the data processing circuit are also configured to implement the following steps:

[0011] - acquisition of at least one surface image of the element;

[0012] - determination of minutiae of an imprint of the element from the image of surface ;

[0013] - characterization of the minutiae determined from the surface image;

[0014] - comparison of the characteristics of the previously determined minutiae with characteristics of theoretical minutiae expected to identify or not the element.

[0015] According to a particular embodiment, the data processing circuit is configured such that the comparison step comprises a score calculation whose value depends on the correlations between, on the one hand, characteristics of the minutiae determined from the surface image and, on the other hand, characteristics of theoretical minutiae expected to identify the element.

[0016] According to a particular embodiment, the ultrasonic transducer circuit, the control circuit and the data processing circuit are also configured to implement a preliminary enrollment step at the end of which minutiae characteristics of the user are recorded.

[0017] According to a particular embodiment, the ultrasonic transducer circuit is configured to acquire depth images of the element by pulsed Doppler type imaging.

[0018] According to a particular embodiment, the depth images correspond to 3D images or to 2D images taken in a plane parallel or perpendicular to the detection surface.

[0019] According to a particular embodiment, the ultrasonic transducer circuit and the control circuit are configured such that ultrasonic waves emitted during the acquisition of depth images reaches a depth of the element greater than or equal to 1 mm relative to the capture surface.

[0020] According to a particular embodiment, the ultrasonic transducer circuit and the control circuit are configured such that the depth images are acquired at a frequency of at least 50 Hz, and / or for a duration of at least 1 second.

[0021] According to a particular embodiment, the data processing circuit is configured to apply, during the acquisition of successive depth images, beam forming type processing to response signals delivered by the ultrasonic transducer circuit.

[0022] According to a particular embodiment, the data processing circuit is configured to carry out the comparison of the depth images with each other by first calculating absolute difference images between immediately consecutive or non-immediately consecutive depth images.

[0023] According to a particular embodiment, the data processing circuit is configured to implement automatic thresholding type processing of the absolute difference images.

[0024] According to a particular embodiment, the data processing circuit is configured to implement a time sliding window type processing from images resulting from the automatic thresholding type processing.

[0025] According to a particular embodiment, the data processing circuit is configured to implement, between the detection of the modification representative of the blood circulation and the determination of the living or non-living character of the element, a determination of a vasculature of the element from the modification representative of the blood circulation.

[0026] According to a particular embodiment, the data processing circuit is configured to determine whether the element is alive or not based on characteristics of the vasculature of the element.

[0027] According to a particular embodiment, the data processing circuit is configured to implement, after determining whether the element is alive or not:

[0028] - a comparison of characteristics of the representative modification of the cir blood culture with expected characteristics recorded during a prior enrollment step, or

[0029] - the application of a machine learning algorithm on characteristics of the representative change in blood circulation to determine if the item is that of an expected user.

[0030] According to a particular embodiment, the ultrasonic transducer circuit, the control circuit and the data processing circuit are configured to implement implements the preliminary enrollment step for different bearing pressures and / or different positions of the element against the capture surface, and to implement the step of acquiring depth images of the element for at least part of the different bearing pressures and / or different positions of the element.

[0031] According to a particular embodiment, the ultrasonic transducer circuit, the control circuit and the data processing circuit are configured to implement, before the acquisition of the depth images, a step of measuring a pressure of the element against the capture surface, and to implement the acquisition of the depth images with parameters chosen as a function of the pressure of the element against the capture surface.

[0032] According to a particular embodiment, the biometric identification and living organism detection device further comprises an element indicating the pressure of the element against the capture surface. Brief description of the drawings

[0033] These characteristics and advantages, as well as others, will be explained in detail in the following description of particular embodiments given without limitation in relation to the attached figures among which:

[0034] - [Fig.l] and [Fig.2] schematically represent an identification device biometric according to a particular embodiment;

[0035] - [Fig.3] represents, in the form of a schematic diagram, steps taken implemented by a biometric identification and living detection device according to a particular embodiment;

[0036] - [Fig.4] schematically illustrates minutiae identified in an image of surface captured by a biometric identification and living detection device according to a particular embodiment;

[0037] - [Fig.5] represents a first schematic example of depth images acquired by a biometric identification and living detection device according to a particular embodiment;

[0038] - [Fig.6] represents a second schematic example of depth images acquired by a biometric identification and living detection device according to a particular embodiment;

[0039] - [Fig.7] represents processing data obtained from images of depth acquired by a biometric identification and living detection device according to a particular embodiment;

[0040] - [Fig.8] represents examples of images calculated during image processing depth acquired by a biometric identification and life detection device according to a particular embodiment. Description of the embodiments

[0041] The same elements have been designated by the same references in the different figures. In particular, the structural and / or functional elements common to the different embodiments may have the same references and may have identical structural, dimensional and material properties.

[0042] For the sake of clarity, only the steps and elements useful for understanding the embodiments described have been shown and are detailed. In particular, different elements (ultrasonic transducer circuit, control circuit, data processing circuit, etc.) and different steps implemented (acquisition of images, processing of acquired images, determination of minutiae, details of calculations performed) are not detailed. Those skilled in the art will be able to implement these elements in detail from the functional description given here.

[0043] Unless otherwise specified, when referring to two elements connected to each other, this means directly connected without intermediate elements other than conductors, and when referring to two elements connected (in English "coupled") to each other, this means that these two elements can be connected or be connected by means of one or more other elements.

[0044] In the following description, when reference is made to absolute position qualifiers, such as the terms "front", "back", "top", "bottom", "left", "right", etc., or relative position qualifiers, such as the terms "above", "below", "upper", "lower", etc., or to orientation qualifiers, such as the terms "horizontal", "vertical", etc., reference is made, unless otherwise specified, to the orientation of the figures in a normal position of use.

[0045] Unless otherwise specified, the expressions "about", "approximately", "substantially", and "of the order of" mean to within 10%, preferably to within 5%.

[0046] Throughout the document, the term “vasculature” is used interchangeably to designate blood vessels or micro-blood vessels, of the artery and / or vein type, and is used as a synonym for the term “microvasculature”.

[0047] A biometric identification and living detection device 100 according to a particular embodiment is described below in connection with FIGS. 1 and 2.

[0048] In the example visible in [Fig.l], the device 100 is configured to implement a biometric identification from an element 101 of the user of the device 100, this element 101 corresponding for example to one or more fingers of this user. The device 100 comprises a capture surface 102 on which this element 101 of the user of the device 100 is intended to be placed during the identification.

[0049] The device 100 comprises at least one ultrasonic transducer circuit 104, a control circuit 106, and a data processing circuit 108. In the schematic example shown in [Fig.2], the control circuit 106 may be electrically coupled to the ultrasonic transducer circuit 104 and thus be able to transmit to the ultrasonic transducer circuit 104 signals controlling the emission of ultrasound for the measurements to be carried out, and also able to receive electrical measurement signals transmitted by the ultrasonic transducer circuit 104. The control circuit 106 may also be coupled to the data processing circuit 108 in order to transmit to the data processing circuit 108 the measurement results obtained from the ultrasonic transducer circuit 104.

[0050] The circuit 104 may comprise a plurality of ultrasonic transducers for carrying out the image acquisitions required for biometric identification and the detection of living beings. The ultrasonic transducers are for example arranged in a matrix manner, or in another manner adapted to the captures to be carried out. The number of ultrasonic transducers of the circuit 104 may depend on the dimensions of the capture surface 102 and / or the desired resolution.

[0051] In the described embodiment, the device 100 may be configured to perform acquisitions of at least one surface image of the user's finger(s) arranged on the capture surface 102 of the device 100 and in contact therewith, i.e. an acquisition of at least one fingerprint image formed of ridges and valleys present on the surface of the skin of the finger(s) arranged on the capture surface 102. In the device 100, the acquisition of a surface image of the user's finger(s) arranged on the capture surface 102 and in contact therewith is based on the fact that the ultrasound emitted by the transducer circuit 104 is reflected more significantly against the air present in the valleys of the fingerprint of the finger(s) than against the ridges of the print.

[0052] In addition, the device 100 may also be configured to perform successive depth image acquisitions of the finger(s) arranged on the capture surface 102 of the device 100 and in contact with it, and to detect, by comparing the depth images with each other, at least one modification representative of blood circulation in a vasculature of this or these fingers. The depth images may correspond to images of at least one part of the finger(s) which is located under the skin of this or these fingers, for example at a depth of between 1 mm and 6 mm (this depth corresponds to the dimension parallel to the Z axis visible in [Fig.l] and which is for example substantially perpendicular to the capture surface 102).

[0053] The ultrasonic transducer circuit 104 is for example configured to acquire depth images of the finger(s) placed on the capture surface 102 by pulsed wave Doppler imaging. In such a configuration, the circuit 104 is configured to emit a series of ultrasonic pulses. The responses obtained in such a configuration do not correspond to a change in the frequency of the emitted wave, but correspond to a pseudo-Doppler effect, i.e. a change in the time interval between the received echoes relative to the time interval between the emitted waves. In the device 100, pulsed Doppler imaging can therefore be used to determine the presence of vascular movements in the acquired depth images, the presence of these movements serving to prove that the element 101 is part of a living being.

[0054] The ultrasonic transducer circuit 104 may for example comprise transducers of the CMUT type (Capacitive Micromachined Ultrasonic Transducer) or of the PMUT type (Piezoelectric Micromachined Ultrasonic Transducer) or of another type.

[0055] In the exemplary embodiment described herein, the control circuit 106 may be configured to provide the ultrasonic transducer circuit 104 with electrical excitation signals causing the circuit 104 to emit ultrasonic waves, and to receive electrical response signals generated by the circuit 104 as a result of receiving ultrasonic waves reflected by the element 101 present on the capture surface 102.

[0056] In the described embodiment, the data processing circuit 108 may be configured to analyze and process the electrical response signals received by the control circuit 106 and sent by the ultrasonic transducer circuit 104. The processing circuit 108 may comprise, for example, at least one microprocessor coupled to at least one memory for processing the received data.

[0057] The control circuit 106 and the data processing circuit 108 can be configured to perform biometric identification and detection of living beings by implementing the steps described below and represented in the form of a diagram in [Fig.3].

[0058] During a first step 202, the device 100 can perform a detection of the presence of an element on the capture surface 102. If an element is actually detected on the capture surface 102, a second step 204 can be implemented. In the absence of a detected element, this detection step 202 can be repeated until an element is detected on the capture surface 102. The detection step 202 can be repeated continuously or at time intervals which may be regular or not.

[0059] The device 100 can then be configured such that it is capable of acquiring at least one surface image of the element 101 present on the capture surface 102 (step 204). For example, this configuration of the device 100 can include sending an acquisition order, or a sequence of instructions, to the control circuit 106 for acquiring the surface image of the element 101.

[0060] During a following step 206, an acquisition of at least one surface image of the element 101 present on the capture surface 102 can then be implemented. This acquisition can include the emission of a series of ultrasonic signals in the form of pulses by the transducer circuit 104, then the reception of the echoes by the transducer circuit 104 and the processing of the responses obtained to obtain the surface image of the element 101, that is to say the image of the user's fingerprint(s).

[0061] The processing of the responses obtained to obtain the surface image of the element 101 corresponds for example to a signal processing of the “beamforming” or “micro-beamforming” type, also called beamforming or spatial filtering. Examples of signal processing of the “beamforming” type are described in the document by MF Rasmussen et al., “3-D imaging using row-column-addressed arrays with integrated apodization - part i: apodization design and line element beamforming”, IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 62, no. 5, pp. 947-958, May 2015, as well as in the document by Perrot Vincent et al., “So You Think You Can DAS? A Viewpoint on Delay-and-Sum Beamforming.” Ultrasonics, vol. 111, March 2021, 106309.Alternatively, it is possible to carry out the transmission and reception of signals from the transducer circuit 104 by multiplexing, to carry out an image acquisition in B mode focused at the surface of the element 101. According to another variant, it is possible to carry out an orthogonal addressing of the transducer matrix of the circuit 104 to emit a series of ultrasonic pulses in the form of focused waves, or plane waves, or to carry out a complete matrix capture of the RCA-OPW type (Row column ad-dressing orthogonal plane wave imaging in English).

[0062] The acquisition of the surface image of the element 101 may also include the implementation of other steps not detailed here: filtering of the response signal obtained, envelope detection, logarithmic compression, etc.

[0063] The acquired surface image may correspond to the capture of a print of the element 101, for example one or more fingerprints when the element 101 corresponds to one or more fingers of the user. The data relating to the acquired image are for example stored in a memory of the device 100 or in an external memory of the device 100 for example connected to the device 100 by a communication link.

[0064] During a following step 208, from the previously acquired surface image, the data processing circuit 108 can determine or extract minutiae of the imprint of the element 101 obtained on this image. This determination of minutiae can be performed by one or more image processing algorithms not described in detail here and known to those skilled in the art.

[0065] In [Fig.4], an example of a surface image acquired by the device 100 is represented and designated by the reference 110. This surface image 110 corresponds here to the imprint of a finger present on the capture surface 102 of the device 100. The minutiae determined in this surface image 110 are represented by points and designated by the reference 112.

[0066] During a following step 210, each of the minutiae extracted from the surface image is characterized (type of minutiae, position in the plane of the surface image, orientation of the minutiae, orientation relative to the other minutiae, etc.).

[0067] During a following step 212, the characteristics of the minutiae are compared with expected characteristics of minutiae, for example previously determined during a preliminary step of enrolling the user, to confirm or not the identity of the user whose finger(s) form the element 101.

[0068] According to a particular example, the comparison of the characteristics of the minutiae 112 with the expected characteristics of minutiae may include a score calculation whose value depends on the correlations between, on the one hand, the characteristics of the minutiae 112 previously determined and, on the other hand, the expected characteristics of minutiae to confirm or not the identity of the user. For example, the score calculation method implemented may make it possible to obtain a score close to or equal to 1 when the user of the device 100 is indeed the person whose identity must be verified, and a score close to 0 in the opposite case.

[0069] According to one example, such a score can be calculated for each previously determined minutia. Then, it is possible to select some or all of the scores obtained and to calculate an overall score representative of the previously calculated scores (sum, average, etc.). The score calculation method is not described in detail here and may correspond to an adaptation of one of the methods known to those skilled in the art and for example applied for identification by comparison of fingerprints.

[0070] The value of the overall score can then be compared with a threshold value in order to evaluate the correspondence between the biometric measurement carried out and the expected biometric data, and thus confirm or not the identity of the user.

[0071] Then, the device 100 can be configured such that it is capable of carrying out successive acquisitions over time of several depth images of the element 101 present on the capture surface 102 (step 214). For example, this configuration of the device 100 can comprise the sending of an acquisition order, or a sequence of instructions, to the control circuit 106 for the acquisition of the depth images of the element 101.

[0072] During a following step 216, an acquisition of several successive depth images of the element 101 present on the capture surface 102 can be implemented. The acquisition of the depth images can be carried out at a frequency for example at least equal to 50 Hz, for example equal to or greater than 150 Hz, and for a duration sufficient for the series of captured depth images to include at least one movement of arterial walls and tissues located around these walls due to the blood circulating in the vasculature of the element 101, for example a duration of between 1 second and 2 seconds, or between 1.5 seconds and 2 seconds. The ultrasonic waves emitted for this acquisition can correspond to plane waves. The captured depth images can correspond to 2D or 3D images (in volume).

[0073] The processing of the responses obtained to obtain the depth images of the element 101 corresponds for example to a processing of “beamforming” type signals.

[0074] The acquisition of the depth images of the element 101 may also include the implementation of other steps not detailed here: filtering of the response signals obtained, envelope detection, logarithmic compression, etc.

[0075] In the device 100, thanks to the series of depth images acquired, movements of the tissues and arterial walls due to blood circulation can be detected and then analyzed to determine proof of the living nature of the element 101 present on the detection surface 102. The data relating to the depth images obtained can be stored in a memory of the device 100 or in an external memory of the device 100, for example connected to the device 100 by a communication link.

[0076] In [Fig. 5], a first schematic example of depth images acquired by the device 100 are represented and are designated by the reference 164. The succession of A depth images 164 acquired at times t0 to TN show the evolution of the movement of the arterial walls of a vessel when blood circulates in this vessel. In this example, the depth images 164, which are 3D images, are represented in [Fig. 5] in the form of 2D images taken in a plane substantially parallel to the capture surface 102.

[0077] In [Fig.6], a second schematic example of depth images 164 acquired by the device 100 are represented. The succession of the A depth images 164 acquired at times t0 to TN show the evolution of the movement of the arterial walls of a vessel, seen in section, when blood circulates in this vessel. In this example, the depth images 164 correspond to 2D images taken in a plane perpendicular to the capture surface 102 and perpendicular to the vessel considered in this acquisition.

[0078] In the schematic examples of Figures 5 and 6, only the movement in the vessel in which the blood circulates is represented. In the actual images acquired, the tissues located around the vessel also undergo movements related to the circulation of blood in the vessel.

[0079] During a following step 218, a detection of the displacements or movements in the acquired depth images is implemented, then an analysis of these movements is then implemented in order to determine whether these movements are linked to the circulation of blood in the vasculature of the element 101, and therefore determine proof of the living nature of the element 101.

[0080] According to a first example, the processing implemented to detect the modifications between the depth images and determine whether these are representative of blood circulation in the vasculature of the element 101 may correspond to the implementation of spatio-temporal filtering, or “clutter filtering” in English. For example, the estimation of the movements linked to the circulation of blood in the vasculature may be carried out by the application of Casorati matrices, then by SVD decomposition (“singular value decomposition” in English) and selection of the components corresponding to the movements of the tissues, and reconstruction of a 3D image, defining voxels of the detected movements. The document by Demené C. et al. “Spatiotemporal Clutter Filtering of Ultrafast Ultrasound Data Highly Increases Doppler and fUltrasound Sensitivity”, IEEE Trans Med Imaging. 2015 Nov, 34(11), 2271-85, describes an example of implementing such spatiotemporal filtering. A maximum intensity projection can then be performed to reduce the amount of data processed and obtain 2D images. A convolutional neural network can then be used to determine whether the identified differences correspond to a representative change in blood flow in the vasculature. This first example can advantageously be implemented when the pressure applied by the element 101 on the capture surface 102 is low and blood flow takes place in the arteries and veins of the element 101.

[0081] According to a second example, the detection of modifications representative of a blood circulation in the vasculature of the element 101 can be carried out by defining vector fields representative of the directions and amplitudes of the differences between the successively acquired depth images. Such detection approaches the so-called “speckle tracking” technique as described in the document by L.N. Bohs et al., “Speckle tracking for multi-dimensional flow estimation,” Ultrasonics, Volume 38, Issues 1-8, 2000, pages 369-375, as well as in the document by P. Joos et al., “High-Frame-Rate Speckle-Tracking Echocardiography,” IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 65, no. 5, pp. 720-728, May 2018, this “speckle tracking” technique being able to be adapted to carry out a such detection. These vector fields can be aggregated and classified to determine whether a change between successively acquired depth images is representative of blood flow in the vasculature. With such detection, the depth image acquisition frequency chosen can be greater than 150 Hz.

[0082] According to a third example, the detection of modifications representative of a blood circulation in the vasculature of the element 101 can be carried out by a comparison, pixel by pixel in the case of 2D depth images, or voxel by voxel (a voxel corresponding to a pixel in a 3D image) in the case of 3D depth images, of the differences between the successively acquired depth images. The identified differences can then be grouped in order to identify a path of a blood pulse in the vasculature. This path can then be analyzed for example by a skeletonization (“skeletonization”, or “thinning” in English) of this path, then an extraction of characteristics of this path (for example the total length, the number of branches, the number of inflection points, etc.), in order to determine whether the identified differences correspond to a modification representative of a blood circulation in the vasculature.Alternatively, when the depth images are 3D images, the analysis of this path can be carried out by implementing a maximum intensity projection (MIP) of the depth images, along the axis corresponding to the depth of the images (dimension perpendicular to the detection surface 102), in order to reduce the amount of data processed and obtain 2D images of the path. A convolutional neural network can then be used with the obtained image in order to determine whether the identified differences correspond to a modification representative of blood circulation in the vasculature. As previously, such detection approaches the so-called “speckle tracking” technique which can be adapted to carry out such detection.

[0083] According to a fourth example, from the depth images obtained, it is possible to calculate the absolute differences, voxel by voxel, between the data of immediately consecutive or non-immediately consecutive depth images (separated in time by one or more depth images from each other). Optionally, it is possible to additionally apply low-pass spatial filtering to the absolute difference images in order to eliminate isolated modifications. It is then possible to calculate either the maximum of the absolute differences at each instant, or a quantile for example of at least approximately 95% to obtain good robustness with respect to the aberrant maxima present in the images, the result obtained corresponding to the curve designated by the reference 150 in [Fig.7]. In [Fig.7], the peaks 152 correspond to the pulses which appear when the blood circulates in the vessels of the element 101. This fourth example can advantageously be implemented when the pressure applied by the element 101 on the capture surface 102 is strong and the blood circulation only takes place in the arteries (the strong pressure applied preventing the blood circulation in the veins which are located at a shorter distance from the skin than the arteries).

[0084] According to an example, the identification of the peaks 152 is obtained by carrying out an automatic thresholding type processing of the values ​​of the quantiles as represented in [Fig.7]. For example, this automatic thresholding can be carried out by calculating a median, called med, of these quantiles, represented by the line 154 in [Fig.7], to choose a threshold representing a high proportion of pixels having been modified between the successive depth images, for example a threshold S98% representing 98% of the quantiles (represented by the line 156 in [Fig.7]), then to determine, from this threshold and the value of the median, a threshold value S (represented by the line 158 in [Fig.7]) of the quantiles beyond which a movement is considered to correspond to a movement of circulation of blood in the vasculature. This threshold value S is for example calculated according to the formula:

[0085] [Math.l] 5 = ax S98% + ( 1 - a ) x med

[0086] with a corresponding to a coefficient whose value can be chosen according to the characteristics of the device 100, and for example equal to 0.7.

[0087] In the example described in connection with [Fig.7], this amounts to identifying the moments, or instants, for which the quantiles exceed the threshold S.

[0088] Furthermore, the threshold S can be used to perform automatic thresholding of the absolute difference images as described in the fourth example above and shown in [Fig.8] (designated by the reference 160). Then, it is possible to apply a temporal sliding window type processing, or "temporal sliding window" in English, in order to obtain an image representing the propagation of the movement. For this, for each voxel or pixel, the following formula can be applied:

[0089] [Math.2] CurrentAWT = fi* PrevAWT + CurrentThresDif f

[0090] with [^corresponds to a coefficient whose value can be chosen according to the characteristics of the device 100, and for example equal to 0.6;

[0091] PrevAWT corresponds to the value of the CurrentAWT parameter calculated for the previous image;

[0092] CurrentThresDiff corresponding to the current absolute difference image, even actually spatially filtered by a low-pass filter, designated by the reference 160 in [Fig.8].

[0093] [Fig. 8] represents examples of images 160 of absolute differences as described above. In this figure, the reference 162 designates the final image obtained representing the propagation of the observed movement. In this example of final image, the brighter a pixel is, the more this pixel has undergone change in the sequence of images 160.

[0094] During a following step 220, it is determined whether the element 101 has a living character or not depending on the detection or not of the modification representative of a blood circulation in the vasculature of the element 101.

[0095] The information obtained from the depth images and relating to the movements observed in these images can be used to determine whether these movements are linked to the circulation of blood in the vasculature of the element 101. For example, it is possible to determine whether the observed movements are linked or not to the circulation of blood in the vasculature of the element 101 depending on the regularity of the captured pulses (corresponding to the peaks 152) and / or the duration between these pulses.

[0096] According to another example, it is possible to use a living / non-living classifier which can correspond to a 2D convolutional neural network (which can use in this case the images after the application of a MIP type process, or “Maximum Intensity Projection” in English, or maximum intensity projection) or 3D.

[0097] According to another example, this living / non-living classification can be carried out from indicators such as the number of voxels crossed by the pulse, the total quantity of energy of the pulse (for example the sum of all the voxels), shape indicators (length / width / depth ratios, number of branches after skeletonization, length of each branch of the skeleton), etc.

[0098] In the examples previously described, the element 101 is determined to have a living character since a modification detected in the depth images is indeed representative of blood circulation in the vasculature of the element 101. Alternatively, it is possible that the data processing circuit 108 is configured to implement, after determining whether or not the element 101 is living, a step making it possible to confirm or not the identity of the user on the basis of the information previously obtained. For example, it is possible that the device 100 is configured such that the shape and / or the position of the propagation of the movement observed in the final image (image 162 of [Fig.8]) also serve to confirm that the element 101 corresponds to at least one finger of a living person, and that this person is the expected user.For this, features such as shape and / or position parameters expected for the representative modification of the blood circulation in the vasculature of the element 101 can be acquired and stored during a prior enrollment step. Then, an alignment of the final image with the surface image can be carried out, then a spatial correlation. between these characteristics of the final image obtained and those obtained during enrollment can be determined. Depending on the level of correlation obtained, it is possible to determine whether the user corresponds to the expected user. According to another example, it is possible to use a machine learning algorithm applied to the characteristics of the modification determined in the depth images, and determine, based for example on a score determined by the algorithm, whether the element 101 belongs to the correct user.

[0099] As a variant of the exemplary embodiment described above, it is possible that steps 208 to 212 related to the processing of the surface image carried out after its acquisition are implemented after the acquisition of the depth images or even after the processing of the depth images. Furthermore, it is also possible that steps 204 and 206 related to the configuration and acquisition of the surface image are implemented after steps 214 and 216 related to the configuration and acquisition of the depth images, or even after step 218 related to the processing of the depth images.

[0100] In the previously described embodiments, the determination of whether the element 101 is alive or not is based on the detection, in the acquired depth images, of at least one movement representative of blood circulation in a vasculature of the element 101. Alternatively, it is possible that the determination of whether the element 101 is alive or not is also based on characteristics of the vasculature of the element 101 determined from the acquired depth images. The vasculature of the element 101 can be determined using the series of acquired depth images and from the movements of the tissues and arterial walls due to the blood circulation which allow a location of the elements of the vasculature, and therefore a determination thereof. Then, characteristics of the vasculature, such as for example the number of vessels identified, their locations, their shapes, etc., are taken into account to determine whether element 101 has a living character or not.

[0101] The device 100 can thus form a complete biometric system integrating functionalities for identifying and detecting living things based on detection of blood circulation in the vasculature. The device 100 can in particular make it possible to have:

[0102] - a better level of security compared to an identification device based only on a fingerprint capture;

[0103] - better level of guarantee of detection of living things due to the reliability of a detection of life based on blood circulation detection.

[0104] In the example described above, the device 100 performs biometric identification by comparing the measured data included in the surface image with expected data corresponding to those of a person for whom the identity is verified. Alternatively, it is possible for the device 100 to perform a comparison of the measured data with expected data corresponding to those of several people, the device 100 being able in this case to verify whether the user corresponds to one of these people.

[0105] In the above description, the device 100 can compare the measurements made with expected data stored in a memory of the device 100. Alternatively, it is possible for the device 100 to connect to a remote database in which the expected data corresponding to those of one or more persons are stored to check whether the user of the device 100 corresponds to this person or one of these persons. It is also possible, in another alternative, to connect the device 100 to a smart card containing the data recorded during a prior enrollment. The data comparison step can then be carried out in the smart card and not in the device 100.

[0106] In the above description, the device 100 is used to directly carry out a confirmation or not of the identity of the user of the device 100. Prior to this identification, the device 100 can be used to carry out an initial acquisition of data (surface image(s) and depth images) during a prior enrollment, which will serve as expected data during future identifications of the user of the device 100. Other uses of the device 100 can also be envisaged.

[0107] As a variant of the example previously described, the device 100 can implement additional steps to those previously described to carry out biometric identification and detection of the living. For example, the device 100 can be used to characterize the surface of the element 101 in order to verify that it corresponds to skin, and / or to detect the presence of bones in the element 101.

[0108] According to another example, it is possible that during the preliminary enrollment, depth images are also acquired and that these images are processed so as to determine a positioning of the vessels, for example in relation to a reference point and an orientation of the user's fingerprint(s). Similar processing of the subsequently acquired depth images and a comparison with the expected positioning of the vessels can then be carried out. Such a possibility makes it possible to reinforce the biometric identification and the detection of the living being carried out, for example with regard to the use of a false finger equipped with the expected fingerprint and one or more dummy vessels (the expected positioning of the vessels cannot be known to the person attempting to usurp the identity), or with regard to the use, on a real finger, of a false layer of material reproducing the expected fingerprint.

[0109] According to another example, it is possible that during enrollment, the user is asked to carry out successive acquisitions of several series of depth images of the element 101 by varying the pressure exerted by the element 101 on the capture surface 102 and / or the position of the element 101 on the capture surface 102. The different series of depth images thus acquired can allow highlighting of different parts of the vasculature of the element 101. For example, when the pressure of the element 101 on the capture surface 102 is low, the depth images acquired can correspond to images representative of veins close to the surface of the element 101.When the pressure of the element 101 on the capture surface 102 is greater, the acquired depth images may correspond to images representative of the arteries located at a greater depth under the surface of the element 101. In this other example, the acquisitions of the depth images implemented after enrollment are also carried out by asking the user to carry out several acquisitions by varying the pressure of pressing the element 101 on the capture surface 102 and / or the position of the element 101 on the capture surface 102.

[0110] When several series of depth images are acquired successively by varying the pressure and / or the position on the capture surface 102 during the enrollment step, it is possible to retain only a part of these series of images, for example those on which sufficient vasculature appears to determine whether the element 101 is alive or not. In this case, the acquisitions implemented after the enrollment can be carried out by asking the user to carry out one or more acquisitions with the pressure and / or the position of the element 101 on the capture surface 102 corresponding to those of the series of images from the enrollment phase which have been retained. [YES] In the example above, the acquisitions carried out with the different pressures and / or the different positions of the element 101 on the capture surface 102 can be implemented with the same acquisition parameters or with different parameters which can depend on the requested and / or detected pressing pressure. For example, in the case of pressing the element 101 on the capture surface 102 with a high pressure, the acquisitions (for enrollment and for the following acquisitions) can be implemented with a high frequency (“frame rate” in English), for example of the order of 150 Hz, and a short acquisition duration, for example less than or equal to 2 seconds, while in the case of pressing with a low pressure, the acquisitions can be implemented with a lower frequency, for example of the order of 50 Hz, and a longer acquisition duration, for example between 3 and 4 seconds.Additionally, the vasculature extraction algorithm and its parameterization can be chosen based on the support pressure of . the element 101 on the capture surface 102, for example according to the first example of algorithm previously described when the pressing pressure is low, and / or according to the fourth example of algorithm previously described when the pressing pressure is high.

[0112] In the examples described above, it is possible for the device 100 to be configured to perform a detection of the pressure of the element 101 on the capture surface 102. This detection of the pressure of the pressure can correspond to a measurement, performed by the data processing circuit 108 from the acquired images, of the position of the interface formed between the support surface 102 and the element 101 along a direction substantially perpendicular to the capture surface 102 and relative to the circuit 104. Different techniques can be implemented to perform this measurement.

[0113] According to a first example, the position of the element 101 / capture surface 102 interface relative to the circuit 104 can be determined by detecting at least one maximum value of response signals from a central part of the circuit 104. Indeed, the bearing surface 102 of the device 100 can be formed by a layer of deformable material under which the ultrasonic transducer circuit 104 is located. In this case, the deformation of the layer forming the bearing surface 102 is more or less significant depending on the pressure applied by the element 101 to the bearing surface 102. The detection of the position of this interface along the axis perpendicular to the capture surface 102 can therefore allow the device 100 to evaluate whether the pressure applied by the element 101 to the capture surface 102 is significant or not.

[0114] According to a second example, the position of the element 101 / capture surface 102 interface relative to the circuit 104 can be determined by applying, to the response signals of the circuit 104, at least one beamforming type processing, then a surface determination algorithm.

[0115] According to another particular embodiment, the pressure applied by the element 101 to the capture surface 102 can be determined by measuring a coupling energy between the element 101 and the circuit 104. Indeed, during the ultrasonic measurements carried out by the circuit 104, the more the element 101 presses on the capture surface 102, the greater the sum of the energies of the response signals to these measurements. By determining this sum of the energies of the response signals delivered by the circuit 104 and comparing it to a certain energy threshold value (the value of which depends in particular on the characteristics of the circuit 104), it is therefore possible to determine whether the pressing of the element 101 on the capture surface 102 corresponds to pressing with high pressure or to pressing with low pressure.

[0116] According to another exemplary embodiment, the pressure applied by the element 101 on the capture surface 102 can be determined by measurements of variations or movements detected in the successive volume images acquired.

[0117] The function of measuring the pressure of the element 101 on the capture surface 102 can be used by the device 100 to carry out, during enrollment and / or during acquisitions subsequent to enrollment, an estimation of the force of pressure on the capture surface 102 in order to choose the most suitable acquisition parameters (capture frequency, capture duration, extraction algorithm) as a function of the pressure of pressure which is measured. In addition, it is possible to combine this functionality with a use of the acquired surface image of the element 101 in order to select a part of the element 101 from which a detection of modification representative of the blood circulation will be implemented.

[0118] For the various examples previously described, the device 100 may comprise an element for indicating, or “feedback”, to the user the pressing pressure that he is currently exerting on the capture surface 102 and / or the required pressing duration and / or the required pressing position. This element may for example provide a visual indication, for example in the form of one or more LEDs displaying a color and / or an intensity representative of the pressing pressure and / or the pressing duration and / or the pressing position. Other types of indication are possible, for example sound, haptic, etc.

[0119] The device 100 described above makes it possible, by using an ultrasonic transducer circuit, to perform improved identification and detection of living beings via detection of blood circulation in the vasculature of at least one finger, this detection of blood circulation being obtained by detection of movement, or pulse, of the arterial wall. The device 100 can make it possible to verify, for example, whether the finger present on the capture surface is a real finger and is indeed that of the expected user, but also to detect whether the user is alive or whether the user's finger has not been cut off.

[0120] Various embodiments and variations have been described. Those skilled in the art will understand that certain features of these various embodiments and variations could be combined, and other variations will occur to those skilled in the art.

[0121] Finally, the practical implementation of the embodiments and variants described is within the reach of those skilled in the art from the functional indications given above.

Claims

Claims

1.

2.

3. Biometric identification and life detection device (100) comprising at least one ultrasonic transducer circuit (104), a control circuit (106) and a data processing circuit (108), configured to implement at least the following steps: acquisition (216) of successive depth images (164) of at least one element (101) intended to be placed against a capture surface (102) of the device (100); detection (218), by comparing the depth images (164) with each other, of at least one modification representative of blood circulation in a vasculature of the element (101); determination (220) of the living or non-living character of the element (101) depending on the detection or not of the modification representative of the blood circulation. Biometric identification and life detection device (100) according to claim 1, wherein the ultrasonic transducer circuit (104), the control circuit (106) and the data processing circuit (108) are also configured to implement the following steps: acquisition (206) of at least one surface image (110) of element (101); determination (208) of minutiae (112) of a fingerprint of the element (101) from the surface image (110); characterization (210) of the minutiae (112) determined from of the surface image (110); comparison (212) of the characteristics of the minutiae (112) previously determined with theoretical minutiae characteristics expected to identify or not the element (101). A biometric identification and life detection device (100) according to claim 2, wherein the data processing circuit (108) is configured such that the comparison step (212) comprises a score calculation whose value depends on the correlations between on the one hand characteristics of the minutiae (112) determined from the image surface (110) and on the other hand theoretical minutiae characteristics expected to identify the element (101).

4. Biometric identification and living detection device (100) according to one of claims 2 or 3, in which the ultrasonic transducer circuit (104), the control circuit (106) and the data processing circuit (108) are also configured to implement a preliminary enrollment step at the end of which minutiae characteristics of the user are recorded.

5. Biometric identification and living detection device (100) according to one of the preceding claims, in which the ultrasonic transducer circuit (104) is configured to carry out the acquisition (216) of the depth images (164) of the element (101) by pulsed Doppler type imaging.

6. Biometric identification and living detection device (100) according to one of the preceding claims, in which the depth images (164) correspond to 3D images or 2D images taken in a plane parallel or perpendicular to the detection surface (102).

7. Biometric identification and living detection device (100) according to one of the preceding claims, in which the ultrasonic transducer circuit (104) and the control circuit (106) are configured such that ultrasonic waves emitted during the acquisition of the depth images (164) reach a depth of the element (101) greater than or equal to 1 mm relative to the capture surface (102).

8. Biometric identification and living detection device (100) according to one of the preceding claims, in which the ultrasonic transducer circuit (104) and the control circuit (106) are configured such that the depth images (164) are acquired at a frequency of at least 50 Hz, and / or for a duration of at least 1 second.

9. Biometric identification and living detection device (100) according to one of the preceding claims, in which the data processing circuit (108) is configured to apply, during the acquisition of successive depth images (164), beam forming type processing to response signals delivered by the ultrasonic transducer circuit (104).

10. Biometric identification and life detection device (100) according to one of the preceding claims, in which the data processing circuit (108) is configured to carry out the comparison of the depth images (164) with each other by first calculating absolute difference images (160) between immediately consecutive or non-immediately consecutive depth images (164).

11. Biometric identification and life detection device (100) according to claim 10, wherein the data processing circuit (108) is configured to implement automatic thresholding type processing of the absolute difference images (160).

12. Biometric identification and living detection device (100) according to claim 11, wherein the data processing circuit (108) is configured to implement a time sliding window type processing from images resulting from the automatic thresholding type processing.

13. Biometric identification and living detection device (100) according to one of the preceding claims, in which the data processing circuit (108) is configured to implement, between the detection of the modification representative of the blood circulation and the determination of the living or non-living character of the element (101), a determination of a vasculature of the element (101) from the modification representative of the blood circulation.

14. Biometric identification and living detection device (100) according to claim 13, in which the data processing circuit (108) is configured to carry out the determination (220) of the living or non-living character of the element (101) as a function of characteristics of the vasculature of the element (101).

15. Biometric identification and living detection device (100) according to one of the preceding claims, in which the data processing circuit (108) is configured to implement, after determining whether the element (101) is alive or not: - a comparison of characteristics of the modification representative of the blood circulation with expected characteristics recorded during a prior enrollment step, or - the application of a machine learning algorithm on characteristics of the modification representative of the blood circulation to determine whether the element (101) is the one of an expected user.

16. A biometric identification and life detection device (100) according to claim 15, wherein the ultrasonic transducer circuit (104), the control circuit (106) and the data processing circuit (108) are configured to implement the prior enrollment step for different pressing pressures and / or different positions of the element (101) against the capture surface (102), and to implement the step of acquiring (216) the depth images (164) of the element (101) for at least a portion of the different pressing pressures and / or different positions of the element (101).

17. Biometric identification and living detection device (100) according to one of the preceding claims, in which the ultrasonic transducer circuit (104), the control circuit (106) and the data processing circuit (108) are configured to implement, before the acquisition (216) of the depth images (164), a step of measuring a pressing pressure of the element (101) against the capture surface (102), and to implement the acquisition (216) of the depth images with parameters chosen as a function of the pressing pressure of the element (101) against the capture surface (102).

18. Biometric identification and living detection device (100) according to one of the preceding claims, further comprising an element for indicating the pressure of the element (101) against the capture surface (102).

Images