Method and device for determining the number of partial palm prints to be acquired to reconstruct a complete palm print, corresponding method and acquisition system

Abstract

The present invention relates to a method for determining the number of partial palm print images to be acquired using a fingerprint acquisition device, said fingerprint acquisition device having an acquisition area smaller than the area of ​​a palm representative of a population of individuals. Figure 1

Description

Title of the invention: Method and device for determining the number of partial palm prints to be acquired to reconstruct an entire palm print, corresponding method and acquisition system. FIELD OF THE INVENTION

[0001] The invention relates to the field of biometric imaging, in particular palm print imaging. STATE OF THE ART

[0002] It is common practice to acquire fingerprints, generally known by the general term "fingerprints," which includes "fingerprints" and "palmar prints." Fingerprints are the patterns formed by the dermatoglyphics of the fingers, and palm prints are the patterns formed by the dermatoglyphics of the palms. Dermatoglyphics are the superficial ridges formed on the palms, soles, and fingertips by the dermal ridges and arranged in lines or spirals. They are unique to each individual. The patterns they form constitute a biometric "identity card" by which an individual can be identified.Fingerprinting is a common practice in various administrative procedures with state institutions and in operations carried out by law enforcement with a suspect or defendant in connection with an offense, misdemeanor, crime, or judicial investigation.

[0003] It is known to acquire a full palm print using devices equipped with an acquisition surface large enough to capture an entire palm of an individual in a single acquisition, regardless of the palm's size. However, such a device is bulky and difficult to transport. In practice, particularly for fieldwork, a smaller, more ergonomic, mobile device is preferred. A common example of a mobile device is one whose acquisition surface dimensions conform to the "FAP 60" standard, namely 76 mm x 81 mm.

[0004] However, the average width of a male individual's palm is 89 mm. Therefore, such a device does not allow for the acquisition of the entire palm print in a single acquisition for the vast majority of individuals.

[0005] One solution consists of acquiring partial palm prints, then reconstructing the entire palm using image processing.

[0006] EP 4 273 815 Al describes a method for reconstructing a palm from these images of partial palm prints showing areas of overlap.

[0007] However, the completeness and accuracy of such a reconstruction depend primarily on the palm coverage that partial print images are able to achieve during their assembly. Yet, it is common for the number of partial print images to be insufficient and / or for these images not to cover all relevant areas of the palm. Faced with this problem, a human operator acquiring partial palm prints of an individual using a small acquisition device for the purpose of reconstructing a complete palm print tends to multiply, often unnecessarily, the number of acquisitions, without any guarantee that the entire palm will actually be covered. This results in a waste of time and a lack of operational efficiency. Description of the invention

[0008] The invention improves the reconstruction of full palm prints from partial palm prints by determining the number of partial palm prints to be acquired to cover the entire palm of a hand. In certain embodiments, the invention also reduces the risk of operator error through a partial print acquisition sequence in which each acquisition corresponds to a specific position of an individual's palm relative to the acquisition surface of the acquisition device.

[0009] According to a first aspect of the invention, a method is provided for determining the number of partial palm print images to be acquired using a fingerprint acquisition device, said fingerprint acquisition device being provided with an acquisition surface whose dimension is smaller than the surface area of ​​a palm representative of a population of individuals, the method comprising the following steps implemented by a computer:

[0010] a) Obtaining an image of an individual's fingerprint acquired using said acquisition device;

[0011] b) Measure, from the image of the fingerprint, at least one geometric dimension of the fingerprint;

[0012] c) Estimate at least one geometric dimension of the individual's palm from at least one geometric dimension of the fingerprint, for example using of one or more preference interpolation functions calculated on statistics of different measures over several individuals;

[0013] d) Determine, from at least one estimated geometric dimension of the individual's palm, a number N of image acquisition(s) Ib ..., IN of the individual's palm to cover the entire surface of the palm, each acquisition corresponding to a position Pb ..., PN of the individual's palm on the acquisition surface of the acquisition device.

[0014] The invention is advantageously complemented by the following features, taken alone or in any technically possible combination thereof:

[0015] - In step a), the fingerprint image includes at least one fingerprint digital of the individual, and in step c) a first geometric dimension of the individual's palm is the width of the palm, said width of the palm is estimated using a first interpolation function from the measured geometric dimension of the fingerprint, and a second geometric dimension of the individual's palm is the length of the palm, said length of the palm is estimated using a second interpolation function from the measured geometric dimension of the fingerprint.

[0016] - The first interpolation function is a width interpolation function of palms of a plurality of individuals as a function of at least one geometric dimension of their fingerprints and the second interpolation function is a palm length interpolation function of a plurality of individuals as a function of a length of their palm.

[0017] - In step b) at least one measured geometric dimension of the fingerprint is chosen from the width, length, area of ​​the fingerprint or their combination.

[0018] - In step a), the fingerprint image includes the interdigital part of the palm of the individual, in step b) the width of the interdigital part is measured and in step c), the width and length of the palm are respectively estimated using a first interpolation function of a palm width and a second interpolation function of a palm length as a function of the measured width of the interdigital part.

[0019] In step b), from the image of the papillary impression taken in step a), at least one distance between two papillary ridges is measured; and in step c), the width and length of the palm are respectively estimated using a first interpolation function of a width and a second function of a length of palm as a function of the distance measured between two papillary ridges.

[0020] Step d) comprises the following substeps:

[0021] - compare the estimated geometric dimension of the palm to the dimensions geometric shapes of the acquisition surface of the acquisition device;

[0022] - determine a position Pb ..., PN for each image acquisition Ib ..., IN of the palm of the individual on the acquisition surface of the acquisition device from the result of the comparison.

[0023] - If the estimated width and estimated length are less than the dimensions geometric shapes of the acquisition surface of the acquisition device, then, the number N of image acquisitions is a single image II associated with a position PI centered on the acquisition surface of the acquisition device.

[0024] - If either the estimated length or the estimated width of the palm is greater than geometric dimensions of the acquisition surface of the acquisition device, then:

[0025] a) with regard to length, the number N of image acquisitions is at least two images II, 12 associated respectively with a first position PI corresponding to an upper part of the palm, preferably to at least the upper half of the palm, and with a second position P2 corresponding to an upper part of the palm, preferably to at least the lower half of the palm; i.e.

[0026] b) with regard to the width, the number N of image acquisitions is at least two images II, 12 associated respectively with a first position PI corresponding to the right half of the palm, and with a second position P2 corresponding to the left half of the palm.

[0027] - If the estimated width and estimated length are greater than the dimensions geometric of the acquisition surface of the acquisition device, then, the number N of image acquisition is at least four images II, 12,13,14, associated respectively with a first position PI corresponding to the upper half of the palm, a second position P2 corresponding to the lower half of the palm, a third position P3 corresponding to the right half of the palm, and a fourth position P4 corresponding to the left half of the palm.

[0028] According to a second aspect, the invention proposes a method for obtaining an image of a fingerprint, comprising the following steps:

[0029] e) acquisition, using a fingerprint acquisition device equipped with an acquisition surface whose dimension is less than the surface of a palm representative of a population of individuals, of a number N of acquisition(s) of L, ..., IN of the palm of the individual to cover the entire surface of the palm, each acquisition corresponding to a position Pb ..., PN of the palm of the individual on the acquisition surface of the acquisition device obtained by means of a the first aspect of the invention;

[0030] f) verification that the acquired images cover the entire palm of the individual;

[0031] g) acquisition of at least one corrective image according to a given position if the acquired images do not cover the entire palm;

[0032] h) reconstruction of the entire palm from the images acquired in step e) and possibly in step g).

[0033] The invention according to the second aspect is advantageously complemented by the following features, taken alone or in any technically possible combination thereof:

[0034] - The verification includes the following steps: reconstructing the palm from acquired images; calculate the width and / or length of the reconstructed palm; compare the calculated width and / or calculated length to a reference value.

[0035] - In step e), the acquired images include at least one palm edge, the verification including the following steps: detect, from the acquired images, the palm edges; verify the presence of all palm edges; at least one corrective image being acquired if a palm edge is missing.

[0036] - The process according to the second aspect further comprises the following steps: calculate the width and / or length from the palm edges; compare the calculated width and / or calculated length to a reference value, with at least one corrective image being acquired based on the result of the comparison.

[0037] The invention according to a third aspect proposes a device for acquiring fingerprints comprising a sensor, the sensor comprising an acquisition area less than the entire surface of the palm for an entire population of individuals, the device comprising a computer configured to implement a method according to the second aspect of the invention. DESCRIPTION OF THE FIGURES

[0038] Fig. 1 is a schematic representation of an example of a mobile fingerprint acquisition device.

[0039] Fig. 2 illustrates imprints classified according to 11 classes Cl-Cll in a known manner.

[0040] Fig. 3 illustrates steps of a method for determining an image of a palm print according to an embodiment of the invention.

[0041] Figure 4 illustrates steps of a method for determining a number of hours to be acquired according to an embodiment of the invention.

[0042] Fig. 5 illustrates a hand position for acquiring an imprint in a first embodiment of the invention.

[0043] Fig. 6 illustrates a hand position for acquiring an imprint in a second embodiment of the invention.

[0044] Fig. 7 illustrates a hand position for acquiring an imprint in a first embodiment of the invention.

[0045] Fig. 8 illustrates the correlation between the width, in pixels, of a fingerprint of an individual's hand with the width, in pixels, of the palm of the individual's hand.

[0046] Fig. 9, Fig. 10, Fig. 11 and Fig. 12 illustrate different hand positions for different palm image acquisitions.

[0047] Throughout the figures, similar elements bear identical references. DETAILED DESCRIPTION OF THE INVENTION

[0048] Figure 1 shows a mobile contact fingerprint acquisition device 1 comprising a sensor 2, a processor 3 configured to perform steps that will be described below, and a memory 4 for storing fingerprint images and code instructions for performing various steps (see below). The processor 3 and the memory 4 are, for example, integrated into a PC, tablet, smartphone, or other system equipped with a user interface. The processor 3 may be independent of the sensor 2 or integrated directly into it.

[0049] The sensor 2 is small in size, ensuring the mobility of the device. It conforms to the FAP 60 standard, and comprises an acquisition area 21 measuring 3 x 3.2 inches, or 76 x 81 mm.

[0050] The acquisition surface 21 of the sensor is generally rectangular. An operator invites an individual to place their hand M in a predetermined position. One or more fingers, the hand, or a part of the hand may be placed depending on the type of fingerprint or palm print that needs to be acquired.

[0051] Fingerprints and palm prints can be classified according to several classes, each class corresponding to a particular atomic zone of the palmar surface of the hand to which they relate. The "palmar surface" of a hand is understood to be the surface of the hand that includes the palm, as opposed to the dorsal surface. The palmar surface includes the palm and all the fingers, that is, the thumb, index, middle, ring, and little fingers, the thenar eminence, the fossa, and the hypothenar eminence. The palm of a hand is the part of the palmar surface between the wrist and the base of the fingers. These definitions correspond to those commonly accepted in anatomy.

[0052] In practice, with reference to [Fig. 2], among the different types of fingerprints, 11 classes C1-C11 can be distinguished, corresponding to different anatomical areas of the palmar surface of the hand and adapted to a very large number of needs and use cases: the complete right hand (C1), the complete left hand (C2), the palm of the right hand (C3), the palm of the left hand (C4), the interdigital area of ​​the right hand (C5), the interdigital area of ​​the left hand (C6), the writer's right palm (C7), the writer's left palm (C8), at least two, preferably three, preferably four fingers of the right hand (C9), at least two, preferably three, preferably four fingers of the left hand (CIO), the thumbs of the right and left hands (Cil).

[0053] The sensor 2, whose acquisition surface dimensions conform to the FAP 60 standard, namely 76 mm x 81 mm, does not allow for the acquisition of a complete palm print in a single acquisition for the majority of individuals. The average width of a male palm is 89 mm. In particular, the fingerprints of classes C3 and C4 shown in [Fig. 2] cannot be acquired in a single acquisition for the majority of individuals.

[0054] As illustrated in [Fig.3], using the acquisition device, a determination (DET step) of a number of partial palm print images to be acquired is implemented, each image corresponding to a position of the individual's palm on the acquisition surface 21 of the sensor, to allow a full acquisition of the individual's palm.

[0055] In addition, an acquisition step (ACQ step) of each image is implemented.

[0056] In addition, a verification step (VER step) is implemented to ensure that the image(s) acquired do indeed cover the entire palm, and if the verification is not conclusive, then corrective acquisitions may be required and carried out (DETS step) or an alert invalidating the acquisitions may be generated (GEN step).

[0057] Fig. 4 illustrates all the steps involved in determining (see the DET step mentioned in Fig. 3) a number N of image acquisitions of the individual's palm to cover the entire surface of the palm, each acquisition corresponding to a position PI, ..., PN of the individual's palm on the acquisition surface 21 of the fingerprint acquisition device 1.

[0058] In addition, it is also a matter of obtaining one or more instructions for positioning the individual's hand on the acquisition surface 21, instructions which may be visual or auditory as presented above.

[0059] Thus, the determination (step DET) of the number N of palm print acquisitions of the individual to cover the entire surface of the palm includes a preliminary step of acquiring (step E0) an initial image (see images 1a, 1b, 1e in Figures 5, 6 and 7 respectively) of the individual's fingerprint. To do this, an individual is asked by an operator to position their hand M on the acquisition surface 21 of the sensor 2. This initial image (1a, 1b, 1e) is received (step E1) by the computer for processing. When designating an image acquisition, it is to understand that each image is associated with a particular position of the individual's hand on the acquisition surface 21 of sensor 2.

[0060] From the initial image la, Ib, le, a geometric dimension of the individual's fingerprint is measured (step E2).

[0061] Then, based on the measured geometric dimension, an estimation (step E3) of at least one geometric dimension of the individual's palm is performed using one or more interpolation functions. A geometric dimension of the individual's palm can be chosen from the length, width, and / or area. For the sake of simplicity, the palm is generally assumed to be rectangular.

[0062] Depending on the geometric dimension of the individual's palm thus estimated, the number N of acquisition(s), and the hand positions corresponding to these N acquisitions, which will cover the entirety of the individual's palm are determined (step E4).

[0063] Each position is stored in memory 4 of a data processing device in the form of visual and / or sound instructions in order to perform the subsequent acquisition of each image (ACQ step). First method of implementation

[0064] According to a first embodiment, with reference to [Fig. 5], the initial image includes at least one fingerprint of the individual, for example, the fingerprints of the four fingers of the left or right hand. In the example illustrated in [Fig. 5], the individual positions their fingers on the acquisition surface 21 of the sensor 2 so that the tips of the four fingers (shown in black) are positioned on the sensor; the acquisition of the initial image (step EOa) is then carried out.

[0065] In this initial image, at least one geometric dimension of one or more fingerprints is measured (step E2). Such a measurement is carried out in pixels or in dimensional units, these two quantities being generally proportional. A geometric dimension of the fingerprint is the width, length, or area of ​​the fingerprint.

[0066] Next, an estimation of the width of the palm is carried out (step E3) from the geometric dimension of the prints measured previously.

[0067] Such an estimation of the width is, for example, carried out from reference measurements giving a geometric dimension of the fingerprints for a palm width of reference individuals, these reference measurements being stored in a memory 4.

[0068] Thus, for a measured geometric dimension of footprints, it is necessary to determine the closest geometric dimension of reference footprints in order to deduce the width of the individual's palm.

[0069] To do this, a first interpolation function fl is obtained from a set of measurements of the palm widths of a plurality of individuals as a function of at least one geometric dimension of their fingerprints.

[0070] The first interpolation function fl is, for example, obtained from a correlation analysis between a geometric dimension of fingerprints and the palm width of a statistical population of reference individuals. The geometric dimension of the fingerprints can be a dimension of a particular finger, or a combination of all the fingers or some of them, with calculation of averages, weighted or unweighted.

[0071] Preferably, a first interpolation function fl of the width A of the palm is used from the width of a fingerprint to obtain the width of the palm. Thus, from the measurement of the width of the individual's fingerprint, an estimate of the width of their palm is obtained (step E3).

[0072] An example of the correlation between the width of fingerprints of different fingers and the width of the palm is illustrated in [Fig. 8]. In this figure, the dimensions are expressed in pixels. The x-axis represents the width of the middle finger, and the y-axis represents the width of the palm for each hand of several individuals. The center line represents the linear relationship between palm width and the width of a fingerprint of a finger (constant finger / palm ratio). The upper and lower lines represent the lower and upper limits of a 95% confidence interval centered on the center line. Almost all the measurement points lie between the upper and lower lines.

[0073] In addition, for any geometric parameter obtained with a first interpolation function fl, it is accompanied by a confidence interval.

[0074] From the width A of the palm, the length L of the palm is determined using a second interpolation function f2 (step E3).

[0075] The second interpolation function f2 is, for example, derived from anthropometric studies indicating, in particular, the 5th and 95th percentiles of palm width and length measurements. An example of such a study is described in NASA MAN-SYSTEMS STANDARDS - 3 ANTHROPOMETRY AND BIOMECHANICS, Revision B, July 1995, Volume 1, Section 3.

[0076] In this example, ,1a second interpolation function f2 allows a correspondence to be established between the width and length of the individual's palm, for example by a linear relationship between palm width and length from the lengths and widths at the 5th and 95th percentiles. Second embodiment

[0077] According to a second embodiment, with reference to [Fig. 6], the initial image Ib of the fingerprint comprises the interdigital portion of the individual's palm. This image can be acquired (step EOb) by positioning the interdigital portion of the individual's hand M along the diagonal of the acquisition surface 21 of the sensor 2 as illustrated in [Fig. 6]. The interdigital portion corresponds to the top of the palm at the base of the fingers.

[0078] For example, with a sensor 2 with a diagonal of 111 mm, it is possible to acquire the width of the palm (measured on the interdigital part) for almost all individuals.

[0079] Indeed, according to certain anthropometric studies, notably the one cited above, 95% of males have a palm width of less than 96 mm and 50% of males less than 89 mm (a difference of 7 mm). A diagonal of 111 mm (therefore 15 mm more than 96 mm) is sufficient to account for a very large majority of individuals.

[0080] From the initial image Ib, the width of the interdigital space is measured (step E2), for example using image processing to measure the edge-to-edge distance of the palm. The width of the interdigital space corresponds approximately to the width of the palm.

[0081] Similar to that of the first embodiment, the length L of the individual's palm is estimated using an interpolation function f3 of a palm length as a function of a width of the interdigital part. Third mode of implementation

[0082] According to a third embodiment, the initial fingerprint image includes papillary ridges. The initial image can be an image of part of the palm and / or fingers, as illustrated in the example in [Fig. 7]. In this example, the initial image can be acquired (step E0) by positioning the individual's hand so that the central part of the palm is in contact with the acquisition surface 21 of the sensor 2. By way of example, [Fig. 7] shows the position of the individual's hand for acquiring (step E0) an initial image showing the central part of the palm (in grayscale).

[0083] From the initial image, at least one distance between two papillary ridges is measured. Alternatively, an average of the widths between papillary ridges can be used for a more representative statistic.

[0084] Then, from the distance measured between two papillary ridges, it is possible to deduce the length and width of the individual's palm using an interpolation function of palm width and length as a function of the distance measured between two papillary ridges. An example of such a function is described in the article Sharma et al. (2022). Is fingerprint ridge density influenced by hand dimensions? Acta Biomed. This function can be advantageously refined using a measurement campaign on a representative population of individuals.

[0085] Determination of the number of acquisitions N (step E4)

[0086] Once the width and length of the palm have been obtained, the next step is to determine the number N of acquisitions required to cover its entire surface, and possibly generate guidance instructions to guide the operator and the individual whose fingerprints are being acquired (step E4). Several cases can be distinguished.

[0087] Case 1) - If the estimated length L and width A of the palm are less than the geometric dimensions of the acquisition surface 21 of the sensor 2 (length of the acquisition surface Lsensor and width of the acquisition surface lsensor) (for example 8.1 cm in length and 7.6 cm in width) then the acquisition number N is a single image II associated with a position PI of the hand centered on the acquisition surface (see [Fig.9]).

[0088] Case 2) - If the estimated length L of the palm is greater than the length LcapteUr of the acquisition surface 21 of the sensor 2 and if the estimated width A of the palm is less than the width of the sensor lcapteural then the number N of image acquisitions is two images II, 12 associated respectively with a first position PI corresponding to an upper part of the palm, preferably at least the upper half of the palm, and with a second position P2 corresponding to an upper part of the palm, preferably at least the lower half of the palm (see [Fig. 10]).

[0089] Case 3) if the estimated length L is less than the length LcapteUrde the acquisition surface 21 of sensor 2 and if the estimated width A of the palm is greater than the width lcapteurde the acquisition surface 21 then the number N of acquisitions is two images II, 12 (see [Fig.1 1]) associated respectively with a first position PI corresponding to the left half of the palm, a second position P2 corresponding to the right half of the palm.

[0090] Case 4) - If the estimated length L of the palm is greater than the length Lsensor of the acquisition surface 21 of sensor 2, and if the estimated width A of the palm is greater than the width of the sensor Lsensor, then the number N of image acquisitions is four images II, 12, 13, 14, associated respectively with a first position PI corresponding to the upper half of the palm, a second position P2 corresponding to the lower half of the palm, a third position P3 corresponding to the right half of the palm, and a fourth position P4 corresponding to the left half of the palm. Therefore, N = 4 acquisitions II, 12, 13, 14 are required in total, one for each of the following palm sectors (see [Fig. 12]): upper left, upper right, lower left, and lower right. It should be noted that different acquisition orders are possible, but it In practice, it is preferable not to start with two sectors located diagonally from each other because the image overlap is less in this case (as such, the overlap allows the palm to be reconstructed from the different acquisitions).

[0091] Depending on the case, 1, 2, or 4 partial images from different positions are required based on estimates of the geometric dimensions of the palm. Depending on the case, a small number of images is required compared to a process where a much larger number would be required to ensure coverage of the entire palm. Obtaining a fingerprint image

[0092] After determining the number of acquisitions necessary to fully cover an individual's palm, the acquisition (ACQ step) is carried out on the acquisition surface 21 of the sensor 2 of each image L, ..., IN according to the position Pb ..., PN associated with it.

[0093] According to certain embodiments, the method further comprises a verification (VER) that the acquired images cover the entire palm, followed by an acquisition (ACQS) of at least one corrective image from a given position if the acquired images do not cover the entire palm. Finally, a reconstruction (REC step) of the entire palm from the acquired images, possibly supplemented by at least one corrective image, is performed.

[0094] The verification step can be implemented in several ways.

[0095] According to a first possibility, the palm is first reconstructed (step E51) from the acquired images, and then the width and / or length of the reconstructed palm are calculated (step E52). Next, the calculated width and / or length is / are compared to one or more reference values ​​(step E53). Then, at least one corrective image is acquired (step ACQS) from a given position based on the result of the comparison.

[0096] The reconstruction of the entire palm from the acquired images can be implemented according to the different methods, in particular according to the method described in document EP 4 273 815 AL Such a reconstruction (step E51) includes registration and fusion steps of the different images (in the case where at least two images for two different positions are required).

[0097] If the width and / or length of the reconstructed palm is / are less than a reference value (a threshold or a range of values), it is likely that the acquired images are of insufficient quality, in particular the hand was not positioned correctly to cover the surface of the palm (the same reasoning can be applied to the length of the hand). A number of corrective acquisitions (ACQS step) is then one or two additional acquisitions to properly cover the entire width and / or one or two additional acquisitions to properly cover the entire width. the entire length of the palm. The acquisition device can be configured to alert the operator using an audible or visual signal.

[0098] According to a second possibility, if the acquired images include at least one palm edge, then a palm edge detection (step E61) is performed on the acquired images in order to detect all the palm boundaries (step E62). If a palm edge is missing, then at least one corrective image is acquired (step ACQS).

[0099] If these borders have been detected, and the width and length (and therefore area) are greater than a reference value, and the shape is consistent with that of a palm, it can be concluded that the entire palm has been captured. Conversely, if one or more borders are missing, and / or the length or width is less than the lower limit of the confidence interval, then at least one corrective image acquisition is performed to detect the corresponding borders.

[0100] In particular, corrective acquisitions consist of positioning the part of the palm corresponding to the missing edge towards the center of the acquisition surface 21. For example, if the left edge of the palm is missing, it is because the left part of the palm has overflowed the left part of the sensor, so the left part of the palm is to be positioned towards the center of the sensor 2, or advantageously between the center and the left of the sensor so that the left edge is in the acquisition surface 21. If the left edge and the top edge are missing, then the top and left part of the palm is to be positioned towards the center or at least (and advantageously) between the center and the top-left quadrant.

[0101] According to a third possibility in addition to the second possibility, the width and / or length from the palm edges are calculated (step E63) and the calculated width and / or length are compared to reference values ​​(step E64). Such reference values ​​are, for example, the lower and upper limits of a confidence interval associated with the interpolation functions (see above).

[0102] If a calculated width or length exceeds reference values, it is likely that the hand was incorrectly positioned angularly on the sensor ("crooked"), as the diagonal measurement was mistakenly taken as a measurement of the palm's width or length. To limit the risk of error during the reconstruction of the full palm print, it is advantageous to generate (GEN step) an alert message for the operator, instructing them to check the hand's position and restart the acquisition.

Claims

Demands

1. A method for determining (DET) the number of partial palm print images to be acquired using a fingerprint acquisition device (1), said fingerprint acquisition device being provided with an acquisition surface (21) whose dimension is smaller than the surface area of ​​a palm representative of a population of individuals, the method comprising the following steps carried out by a computer (3): a) Obtaining (E1) an image of a fingerprint of an individual acquired using said acquisition device (1); b) Measuring (E2), from the fingerprint image, at least one geometric dimension of the fingerprint; c) Estimating (E3) at least one geometric dimension of the individual's palm from at least one geometric dimension of the fingerprint;d) Determine (E4), from at least one estimated geometric dimension of the individual's palm, a number N of image acquisition(s) Ib ..., IN of the individual's palm to cover the entire surface of the palm, each acquisition corresponding to a position Pb ..., PN of the individual's palm on the acquisition surface (21) of the acquisition device (2).;

2. A method according to claim 1, wherein the fingerprint image comprises at least one fingerprint of the individual, and in step c) a first geometric dimension of the individual's palm is the palm width, said palm width is estimated using a first interpolation function (fl) from the measured geometric dimension of the fingerprint, and a second geometric dimension of the individual's palm is the palm length, said palm length is estimated using a second interpolation function (f2) from the measured geometric dimension of the fingerprint.

3. A method according to claim 2, wherein the first interpolation function is a palm width interpolation function (fl) of a plurality of individuals as a function of at least one geometric dimension of their fingerprints, and the second interpolation function (f2) is an interpolation function of palm length of a plurality of individuals as a function of the length of their palm.

4. A method according to any one of claims 2 to 3, wherein at least one geometric dimension of the fingerprint is chosen from the width (A), length (L), area (S) of the fingerprint or a combination thereof.

5. A method according to claim 1, wherein the fingerprint image includes the interdigital part of the individual's palm, and in step b) the width of the interdigital part is measured and in step c), the width and length of the palm are estimated respectively using a first interpolation function of a palm width and a second interpolation function of a palm length as a function of the measured width of the interdigital part.

6. A method according to claim 1, such that, in step b), from the image of the fingerprint taken in step a) at least one distance between two papillary ridges is measured; and in step c), the width and length of the palm are respectively estimated using a first function (f3) of interpolation of a width and a second function (f4) of a palm length as a function of the distance measured between two papillary ridges.

7. A method according to any one of claims 1 to 6, wherein step d) comprises the following substeps: - comparing at least one estimated geometric dimension of the palm to the geometric dimensions of the acquisition surface of the acquisition device; - determining a position PB ..., PN for each image acquisition Ii, ..., IN of the individual's palm on the acquisition surface of the acquisition device from the result of the comparison.

8. Method according to claim 7, such that if the estimated width and estimated length are less than the geometric dimensions of the acquisition surface (21) of the acquisition device (2), then the number N of image acquisitions is a single image II associated with a position PI centered on the acquisition surface of the acquisition device.

9. A method according to claim 7, such that if either the estimated length or the width of the palm is greater than the dimensions geometric of the acquisition surface (21) of the acquisition device (2), then - with regard to the length, the number N of image acquisitions is at least two images II, 12 associated respectively with a first position PI corresponding to an upper part of the palm, preferably at least the upper half of the palm, and with a second position P2 corresponding to an upper part of the palm, preferably at least the lower half of the palm; or - with regard to the width, the number N of image acquisitions is at least two images II, 12 associated respectively with a first position PI corresponding to the right half of the palm, and with a second position P2 corresponding to the left half of the palm.

10. A method according to claim 7, such that if the estimated width and estimated length are greater than the geometric dimensions of the acquisition surface (21) of the acquisition device, then the number N of image acquisitions is at least four images II, 12, 13, 14, associated respectively with a first position PI corresponding to the upper half of the palm, a second position P2 corresponding to the lower half of the palm, a third position P3 corresponding to the right half of the palm, and a fourth position P4 corresponding to the left half of the palm.

11. A method for obtaining an image of a whole palm fingerprint, comprising the following steps: e) acquisition (ACQ), using a fingerprint acquisition device (1) having an acquisition surface (21) whose dimension is smaller than the surface area of ​​a palm representative of a population of individuals, of a number N of acquisition(s) of the individual's palm to cover the entire surface of the palm obtained by means of a method according to any one of claims 1 to 10, each acquisition corresponding to a position Pb ..., PN of the individual's palm on the acquisition surface (21) of the acquisition device (2); f) verification (VER) that the acquired images cover the entire palm of the individual; g) acquisition (ACQS) of at least one corrective image according to a given position if the acquired images do not cover the entire palm; h) reconstruction (REC) of the entire palm from the images acquired in step e) and possibly in step g).

12. A method according to claim 11, wherein the verification (VER) comprises the following steps: - reconstruct (E51) the palm from the acquired images; - calculate (E52) the width and / or length of the reconstructed palm; - compare (E53) the calculated width and / or calculated length to a reference value.

13. A method according to claim 11, wherein, in step e), the acquired images include at least one palm edge, the verification (VER) comprising the following steps: - detect (E61), from the acquired images, the palm edges; - verify (E62) the presence of all the palm edges; at least one corrective image being acquired if a palm edge is missing.

14. A method according to claim 13, further comprising the following steps: - calculate (E63) the width and / or length from the palm edges; - compare (E64) the calculated width and / or calculated length to at least one reference value, at least one corrective image being acquired based on the result of the comparison.

15. A fingerprint acquisition device (1) comprising a sensor (2), the sensor (2) comprising an acquisition area (21) smaller than the entire surface of the palm for an entire population of individuals, the device (1) comprising a computer (3) configured to implement a method according to any one of claims 1 to 14.

Images