WO2014198847A1 - Decision device for deciding whether a surface is living tissue or not - Google Patents

Decision device for deciding whether a surface is living tissue or not Download PDF

Info

Publication number
WO2014198847A1
WO2014198847A1 PCT/EP2014/062277 EP2014062277W WO2014198847A1 WO 2014198847 A1 WO2014198847 A1 WO 2014198847A1 EP 2014062277 W EP2014062277 W EP 2014062277W WO 2014198847 A1 WO2014198847 A1 WO 2014198847A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
intensities
standard deviation
surface
perpendicular
Prior art date
Application number
PCT/EP2014/062277
Other languages
French (fr)
Inventor
Alain Thiebot
Andrey Tarasishin
Benjamin PEYRONNEAUD
Florence Guillemot
Aldo MAALOUF
Original Assignee
Morpho
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to FR1355428A priority Critical patent/FR3007170B1/en
Priority to FR1355428 priority
Application filed by Morpho filed Critical Morpho
Publication of WO2014198847A1 publication Critical patent/WO2014198847A1/en

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K9/00Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
    • G06K9/00885Biometric patterns not provided for under G06K9/00006, G06K9/00154, G06K9/00335, G06K9/00362, G06K9/00597; Biometric specific functions not specific to the kind of biometric
    • G06K9/00899Spoof detection
    • G06K9/00906Detection of body part being alive
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K9/00Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
    • G06K9/00221Acquiring or recognising human faces, facial parts, facial sketches, facial expressions
    • G06K9/00288Classification, e.g. identification
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K9/00Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
    • G06K9/20Image acquisition
    • G06K9/2018Identifying/ignoring parts by sensing at different wavelengths
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K9/00Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
    • G06K9/20Image acquisition
    • G06K9/2036Special illumination such as grating, reflections, deflections, e.g. for characters with relief

Abstract

The invention relates to a decision device (100) for deciding whether a surface (50) is a living tissue or not, said decision device (100) comprising: lighting means (102) which emit an incident coherent stream (104) that is polarized along a polarization direction, has a coherence length no lower than 1 cm and is intended to illuminate said surface (50); a sensing means (106, 108) having a polarizing filter perpendicular and/or parallel to the polarization direction and intended for acquiring at least one image of the illuminated surface (50) having stippling; calculation means (112, 113) for calculating, from the perpendicular image and/or from the parallel image, the value of at least one criterion representing the stippling of said image(s); a database containing, for the or each criterion, an range of reference values of said criterion which represents the stippling of a living tissue; comparison means (114) for comparing the or each value thus calculated to the corresponding range of reference values and for outputting information representing said comparisons; and decision means (116) for deciding whether the surface (50) is a living tissue or not, depending on the information output by the comparison means (114).

Description

 Decision device provided for deciding whether a surface is living tissue or not

The present invention relates to a decision device provided for deciding whether a surface is a living tissue or not, and a decision process implemented by such a decision device.

 It finds application in the field of biometric recognition and in particular in the field of identification by recognition of a living tissue such as an iris, a retina, the skin of a face or a hand.

 The identification of a person by the biometric recognition of a face, the iris or a fingerprint or palmar, is used to secure facilities such as buildings or machines. This technology eliminates access codes or cards that can be stolen or falsified. The use of this technology enhances security as the probability of two people having two identical faces is very low.

The biometric recognition is based on comparing the image of a user's biometry with a set of reference images stored in a database. Although piracy of biometric recognition devices is difficult, it is not impossible.

 Indeed, some pirates realize lures, such as a photo or a mask on which is reproduced the biometry to imitate. The pirate can then place the decoy in front of the acquisition device of the biometric recognition device which is then deceived.

 Numerous devices and processes have been implemented to limit the risks of piracy.

 To differentiate a true biometry from a false, the documents US 2008/0025579, WO 2008/108871 and US2008 / 0219522 propose to analyze multispectral images obtained by capturing the specular reflection of a homogeneous light flux on the surface of the skin. or decoy material.

 Document US 2011/0163163 proposes to analyze multispectral images obtained by capturing the diffusion of a homogeneous light flux on the surface of the skin or of the decoy material.

 These solutions have the disadvantage of multiplying the number of images to be acquired and processed.

 An object of the present invention is to provide a decision device for deciding whether a surface is a living tissue or not and that does not have the disadvantages of the prior art.

 For this purpose, a decision device is provided for deciding whether a surface is a living tissue or not, said decision device comprising:

 lighting means emitting a coherent and polarized incident flux in a polarization direction having a coherence length greater than or equal to 1 cm and intended to illuminate said surface;

 a capture means having a polarizing filter perpendicular and / or parallel to the direction of polarization and intended for acquiring at least one image of the illuminated surface presenting scab,

 calculating means provided for calculating, from the perpendicular image and / or the parallel image, the value of at least one criterion representative of the scabs of said image (s),

a database containing for the or each criterion a range of reference values of said criterion which is representative of scabs of a living tissue, comparison means provided for comparing the or each value thus calculated with the corresponding reference value interval and for providing information representative of these comparisons, and

 decision means provided for deciding whether the surface is a living tissue or not depending on the information delivered by the comparison means.

 Advantageously, the representative criteria for scab are chosen from the standard deviation of the intensities of the perpendicular image, the standard deviation of the intensities of the parallel image, a logical and / or arithmetic combination between the standard deviation of the intensities of the parallel image and the standard deviation of the intensities of the perpendicular image, the standard deviation of the intensities of the perpendicular image on which a high pass type frequency filter is previously applied, the difference between type of the intensities of the parallel image on which a high-pass type frequency filter is previously applied, a logical and / or arithmetic combination between the standard deviation of the intensities of the parallel image on which a frequency filter is applied beforehand of the high-pass type and the standard deviation of the intensities of the perpendicular image on which a high pass type frequency filter is previously applied.

 Advantageously, the criteria adopted are the standard deviation of the intensities of the perpendicular image, the standard deviation of the intensities of the parallel image, and the ratio between the standard deviation of the intensities of the parallel image and the standard deviation of the intensities of the perpendicular image.

 Advantageously, the criteria used are the standard deviation of the intensities of the perpendicular image on which a highpass type frequency filter is previously applied, the standard deviation of the intensities of the parallel image on which a prior high-pass type frequency filter, and the ratio between the standard deviation of the intensities of the parallel image on which a high-pass type frequency filter is previously applied and the standard deviation of the intensities of the perpendicular image on which a high pass type frequency filter is applied beforehand.

 Advantageously, the representative criteria for scab are characteristic dimensions of the scab observed and calculated by an autocorrelation calculation.

 Advantageously, the representative criteria for scab is the standard deviation of the power spectral density.

Advantageously, the comparison means comprise the use of a learning mechanism of the ACP, SVM or neuron network type driven on the basis of real tissue and fraud. The invention also proposes a decision method intended to decide whether a surface is a living tissue or not and implemented by a decision device according to one of the preceding variants, the decision method comprising:

 a lighting step during which the lighting means illuminate the surface to be analyzed,

 a capture step during which the perpendicular capture means captures a perpendicular image of the illuminated surface and / or a capture step during which the parallel capture means captures a parallel image of the illuminated surface,

 a calculation step during which the calculation means calculate the value of each of said criteria selected from each captured image,

 a comparison step during which the comparison means compare each value thus calculated with the corresponding reference value interval and deliver information representative of this comparison,

 a decision step during which the decision means decide whether the surface is a living tissue or not depending on the information delivered by the comparison means.

 The characteristics of the invention mentioned above, as well as others, will appear more clearly on reading the following description of an exemplary embodiment, said description being given in relation to the attached drawings, among which:

 FIG. 1 represents a schematic view of a decision device according to the invention,

 FIG. 2 is an algorithm of a decision method according to the invention, FIG. 3 shows the application to a non-living tissue,

 FIG. 4 shows the application to living tissue, and

 FIG. Figure 5 shows an example of test results.

Fig. 1 shows a decision device 100 for deciding whether a surface 50, here a face 50 of an individual is a living tissue or not. In the remainder of the description, the surface 50 is assimilated to a face, but the invention applies in the same way to all living tissues that can be obtained while respecting the integrity of the human body such as the skin of the face, the skin of a finger, the skin of a palm, or such as the iris or the retina. The decision device 100 comprises lighting means 102 provided for illuminating the surface 50 with a coherent incident flux 104 having a linear polarization. The lighting means 102 are chosen for their coherence length greater than or equal to 1 cm so that the latter is sufficient to generate scabs (or interference or speckles) on the illuminated surface 50.

 Preferably, monomode laser diodes having a coherence length well adapted to the application and widely available will be used. On the other hand, all light sources characterized by a coherence length of the order of a few tens of micrometers (such as LEDs) or of the order of a few micrometers (such as white light) or generating no interference is proscribed for the implementation of the present invention.

 The speckle is an interference due to the difference in step (phase shift) of the reflected beams of the coherent source scattered by an irregular reflecting surface (surface scattering). The level of speckle is important for cases of fraud materials with a high surface diffusion (paper, plasters, resins, etc ...), and decreases in the case of the skin because the beams transmitted in the skin by diffusion of volume lose their coherence after multiple internal reflections and thus generate less interference, so a lower speckle level. Some materials have a weak surface diffusion (low tinted silicone) and will have a lower speckle level than the skin.

 Fig. 3 shows in the case of a fraud 301 that the speckle level is important when said fraud 301 is illuminated by an incident beam 302 due to a large surface scattering and represented by the beam 303 and a low volume diffusion represented by the beam 304.

 Fig. 4 shows in the case of a living tissue 401 that the level of speckle is lower when the tissue 401 is illuminated by an incident beam 402 due to a reduced surface diffusion and represented by the beam 403 and a strong diffusion of volume represented by beam 404.

 The principle of the invention is based on the fact that, depending on the absorption rate of a surface illuminated by the incident coherent flux 104, a level of scab ("speckles" in English) due to the roughness of the surface is generated.

The present invention does not seek to suppress the speckles contrary to prior art methods which aim to reconstruct either a faithful image of the object or a target extraction, but it relies on these speckles to characterize the nature of the surface to be studied.

 In particular, the methods currently used for acquiring images with laser illumination seek to limit the generation of scabbing by favoring laser sources with short coherence lengths such as VCSEL lasers or by modulating the laser diodes.

 The documents http://www.princetonoptronics.com/pdfs/7952-15.pdf and http: //www.princetonoptronics.corn/technology/technology.php#l 02 describe that under certain conditions, lasers can illuminate without generating speckles.

 The incident flux 104 may emit at a wavelength of between 550 and

950 nm when the incident flux 104 illuminates the tissues of an individual.

 Advantageously, when the object to be illuminated is the face, the wavelength of 750 to 950 nm will be chosen to avoid dazzling the individual. Preferably, the wavelength chosen will be approximately 785 nm which is the optimal wavelength for the absorption of the skin of the face, iris or retina so as to obtain a minimum of speckle on the skin. living tissue and a maximum of speckle on other materials.

 According to another embodiment, when the object to be acquired is a part of the hand (one or more fingers, the palm ...), we can also use a red laser emitting between 635 and 650 nm. In this case, the system does not light the eye of the user, the latter is not dazzled.

 As explained below, according to the criterion chosen, the decision device 100 also comprises one or two capture means 106 and 108 sensitive to the wavelength chosen according to the object to be illuminated, each being designed to capture a image of the surface 50 illuminated by the coherent stream 104 characterized by a coherence length of at least 1 cm generating scabs on the surface 50. The capture means 106 and 108 here take the form of two cameras each provided with a interference filter for the wavelength of the coherent flux 104.

The perpendicular capture means 106 has a polarizing filter perpendicular to the polarization direction of the coherent stream 104 and the parallel capture means 108 has a polarizing filter parallel to the polarization direction of the coherent stream 104. The image captured by the perpendicular capture means 106 is called the perpendicular image, and the image captured by the parallel capture means 108 is called the parallel image.

 The illumination means 102 and each capture means 106, 108 are oriented towards the surface 50 to be analyzed.

 Depending on whether the surface 50 is a living tissue or a decoy, the scab intensity will be different.

 If the surface 50 is a living tissue, the latter diffuses in volume a significant part of the incident flux (at the wavelengths considered, the penetration of light into the skin is of the order of 1 mm), so the intensity scab is weak. If the surface 50 is a decoy, the material diffuses weakly in volume and strongly on the surface and the scab intensity is higher.

 The principle of the invention is based on the quantification of speckles. This quantification can be done according to several image processing known to those skilled in the art such as for example: frequency analysis, variance, standard deviation.

 In a preferred embodiment, intensity standard deviation analysis is used on all or part of the illuminated surface where living tissue is to be found.

 The standard deviation of the intensities of the perpendicular image, and / or the standard deviation of the intensities of the parallel image, and / or the logical and / or arithmetic combination between the standard deviation of the intensities of the parallel image and the standard deviation of the intensities of the perpendicular image constitute different criteria which after analysis make it possible to decide on the veracity of the surface 50.

 It is also possible to use the standard deviation of the intensities of the perpendicular image on which a high pass type frequency filter is previously applied, and / or the standard deviation of the intensities of the parallel image on which a high pass type frequency filter is used beforehand, and / or the logical and / or arithmetic combination between the standard deviation of the intensities of the parallel image on which a highpass type frequency filter is first applied and the standard deviation of the intensities of the perpendicular image on which a high pass type frequency filter is previously applied.

It is thus possible by analyzing one of these criteria against a range of permissible reference values for living tissue to decide whether the surface 50 is living tissue or not. For example, if the standard deviation of the intensities of the perpendicular or parallel image is outside the expected range, this detects a too reflective fraud with respect to a living tissue.

 Moreover, if the two standard deviations of the intensities are in the reference value ranges, the logical and / or arithmetic combination of the standard deviations of the intensities makes it possible to confirm the coherence between the two standard deviations of the intensities.

 In a particular embodiment, the ratio of the standard deviation of the intensities of the parallel image to the standard deviation of the intensities of the perpendicular image must be greater than one. In fact, the standard deviation of the intensities of the perpendicular image must be smaller than that of the parallel image because the specular reflection which does not lose its coherence is suppressed on the perpendicular image.

 To increase the reliability of the decision device 100, it is possible to perform an analysis of two or more of these criteria.

 The standard deviation of the intensities of an image is calculated by the following operations: - identification of the zones to be analyzed (supposed presence of skin) by detection of the illuminated face on said image, and

 - calculation of the standard deviation of the intensities in each zone.

 In the case of a standard deviation of the intensities of an image on which a high pass type frequency filter is previously applied, a filtering of the image is carried out between the registration and the calculation.

 The decision device 100 also comprises calculation means 112, 113 intended to carry out the preceding operations and thus to calculate the value or values of at least one criterion representative of the scabs of the perpendicular image and / or the filtered parallel image or no.

 The decision device 100 includes a database containing for each of the criteria selected the range of reference values of said criterion which is representative of scabs of a living tissue.

 According to a first preferred embodiment of the invention, the criteria used are the standard deviation of the intensities of the perpendicular image, the standard deviation of the intensities of the parallel image, and the relationship between the deviation type of the intensities of the parallel image and the standard deviation of the intensities of the perpendicular image.

According to a second preferred embodiment of the invention, the criteria used are the standard deviation of the intensities of the perpendicular image on which a high pass type frequency filter is previously applied, the standard deviation of the intensities of the parallel image on which a high-pass type frequency filter is previously applied, and the ratio between the standard deviation of the intensities of the parallel image on which a high-pass frequency the standard deviation of the intensities of the perpendicular image on which a highpass type frequency filter is previously applied.

 It is possible to combine the two preferred embodiments by analyzing standard deviations of the intensities of the original images and the standard deviations of the intensities of the filtered images.

 For each criterion selected for the analysis, the decision device 100 is subjected to a learning phase, during which real living tissues are presented and during which the value of said criterion is recorded. This learning phase makes it possible to collect for each criterion selected, values of said criterion which are representative of a living tissue and thus to create an interval of reference values which, for said criterion, corresponds to a living tissue. It is also possible to present lures to refine the interval. This learning phase is performed once at the start of the decision device 100.

 The decision device 100 further comprises:

 comparison means 114 provided for comparing the value of the or each criterion with the corresponding reference value interval and delivering representative information for each comparison, and

 decision means 116 provided for deciding whether the surface 50 is a living tissue or not depending on the information delivered by the comparison means 114.

 If the criterion adopted is the standard deviation of the intensities of the perpendicular image, the decision device 100 comprises the perpendicular capture means 106 and the calculation means 112 are provided for calculating the standard deviation of the intensities of said image. perpendicular.

 If the criterion adopted is the standard deviation of the intensities of the parallel image, the decision device 100 comprises the parallel capture means 108 and the calculation means 113 are designed to calculate the standard deviation of the intensities of said image. parallel.

If the criterion adopted is a logical and / or arithmetic combination between the standard deviation of the intensities of the perpendicular image and the standard deviation of the intensities of the parallel image, the decision device 100 comprises the perpendicular capture means 106 and the parallel capture means 108 and the calculation means are provided for calculating the standard deviation of the intensities of said perpendicular image, to calculate the standard deviation of the intensities of said parallel image, and calculating the logical and / or arithmetic combination between the standard deviation of the intensities of the perpendicular image and the standard deviation of the intensities of the parallel image.

 If the criterion adopted is the standard deviation of the intensities of the filtered perpendicular image, the decision device 100 comprises the perpendicular capture means 106 and the calculation means 112 are provided for filtering the perpendicular image and calculating the difference -type of the intensities of said filtered perpendicular image.

 If the criterion adopted is the standard deviation of the intensities of the filtered parallel image, the decision device 100 comprises the parallel capture means 108 and the calculation means 113 are provided for filtering the parallel image and calculating the difference -type of the intensities of said filtered parallel image.

 If the criterion adopted is a logical and / or arithmetical combination between the standard deviation of the intensities of the filtered perpendicular image and the standard deviation of the intensities of the filtered parallel image, the decision device 100 comprises the means of perpendicular capture 106 and the parallel capture means 108 and the computing means are provided for filtering the parallel and perpendicular images and calculating the standard deviation of the intensities of said filtered perpendicular image, for calculating the standard deviation of the intensities of said filtered parallel image, and calculating the logical and / or arithmetic combination between the standard deviation of the intensities of the filtered perpendicular image and the standard deviation of the intensities of the filtered parallel image.

 Thus, if the value of each criterion retained is in the range of the reference values, the surface 50 is then considered as a living tissue and is not considered a living tissue if it is outside this range of values of reference.

 Fig. 2 shows a flowchart of a decision method 300 implemented by the decision device 100 and provided to decide whether a surface 50 is a living tissue or not.

 The decision process 300 comprises:

 a lighting step 302 during which the coherent lighting means

102 illuminate the surface 50 to be analyzed,

 according to the criterion chosen,

a capture step 304 during which the perpendicular capture means 106 captures a perpendicular image of the surface 50 thus illuminated and / or a capture step 306 in which the parallel capture means 108 captures a parallel image of the surface 50 thus illuminated,

 a calculation step 306 during which the calculation means 112 calculate the value of each of said criteria selected from each image thus captured and possibly previously filtered by a high-pass type frequency filter,

 a comparison step 308 during which the comparison means 114 compare each value thus calculated with the corresponding reference value interval, and deliver information representative of this comparison,

 a decision step 310 during which the decision means 116 decide whether the surface 50 is a living tissue or not according to the information delivered by the comparison means 114.

 According to a particular embodiment of the invention, the decision process

300 comprises, prior to the lighting step 302, a learning step 312 which consists of a capture of live tissue images by each appropriate capture means 106, 108 according to the criterion (s) retained , a calculation of the value of each criterion retained by the computing means 112, 113 from the image or each image thus captured, and the definition for the or each criterion retained of an interval of the reference values for which the values of said criterion are considered representative of a living tissue. As mentioned above, the learning step 310 is performed only at the start of the decision device 100 and the decision step 310 then loop on the lighting step 302.

 In another embodiment, certain features of the speckle image are analyzed for living and non-living tissue. This method is based on a statistical approach.

 Any statistical calculation assumes that a set of events (criteria) on which the calculations are based is defined. Thus, the statistical set corresponds to the set of possible draws, supposed infinite.

 It is then possible to calculate, for example, the probability of seeing a particular element (a living or non-living tissue) occurring in this set.

 Two statistical criteria are thus defined to define a final score. If this final score is greater than a threshold, the tissue is considered alive. Otherwise, the tissue is considered non-alive.

The statistical criteria are defined based on the physical properties of the speckle that are measured based on the image processing. The criteria are defined as follows:

 Autocorrelation criterion for speckle characterization:

 2nd order statistics describe relative light intensity variations at two points. As such, they give an indication of the grain size in the speckle figure and their distribution. The most commonly used 2nd order statistical parameter in the speckle study is the autocorrelation function.

 The offset for which the autocorrelation function drops to a defined value

(1/2, 1 / e or 1 / e 2 , or pass through 0) can be used to measure a characteristic dimension of the observed speckle, in particular the average size of speckles grains.

 This shift varies from one tissue to another and makes it possible to distinguish between living and non-living tissues.

 Criterion based on the Fourier transform:

 Another second-order statistical parameter employed is the Wiener spectral power density or spectrum which describes the distribution of the spot sizes in the speckle figure. It is given by the Fourier transform of the autocorrelation function of the luminous intensity.

 The second criterion is thus defined by the standard deviation of the spectral density.

 Fig. 5 shows the performance of these two criteria. The points to the left of the line 502 correspond to the non-living tissues while the points to the right of the line 502 correspond to the living tissues.

 The abscissa corresponds to the Fourier score and therefore to the standard deviation of the spectral density, the ordinate corresponds to the speckle score and therefore to the autocorrelation score.

 The final score for classifying tissues between living and non-living is obtained by conducting a principal component analysis to obtain the equation of line 502 that separates the two classes.

 Any score above the threshold defined by the line corresponds to a living tissue.

All of the disclosed mechanisms include the use of an ACP, SVM or neural network-based learning mechanism driven on true tissue and fraud bases.

 Of course, the present invention is not limited to the examples and embodiments described and shown, but it is capable of many variants accessible to those skilled in the art.

Claims

 CLAIMS
1) Decision device (100) for deciding whether a surface (50) is a living tissue or not, said decision device (100) comprising:
 lighting means (102) emitting a coherent and polarized incident flux (104) in a polarization direction with a coherence length greater than or equal to 1 cm and intended to illuminate said surface (50),
 a capture means (106, 108) having a polarizing filter perpendicular to and / or parallel to the direction of polarization and intended for acquiring at least one image of the illuminated surface (50) presenting with scab,
 calculating means (112, 113) for calculating, from the perpendicular image and / or the parallel image, the value of at least one criterion representative of the scabs of said image (s),
 a database containing for the or each criterion a range of reference values of said criterion which is representative of scabs of a living tissue,
comparison means (114) provided for comparing the or each value thus calculated with the corresponding reference value interval and for providing information representative of these comparisons, and
 decision means (116) provided for deciding whether the surface (50) is a living tissue or not depending on the information delivered by the comparison means
(114).
2) Decision device (100) according to claim 1, characterized in that the criteria representative of scab are selected from the standard deviation of the intensities of the perpendicular image, the standard deviation of the intensities of the parallel image , a logical and / or arithmetic combination between the standard deviation of the intensities of the parallel image and the standard deviation of the intensities of the perpendicular image, the standard deviation of the intensities of the perpendicular image on which one applies beforehand a high-pass type frequency filter, the standard deviation of the intensities of the parallel image on which a high-pass type frequency filter is previously applied, a logical and / or arithmetic combination between the standard deviation intensities of the parallel image on which a high pass type frequency filter is previously applied and the standard deviation of the intensities of the perpendicular image on which a frequency filter is applied beforehand el type high pass. 3) Decision device (100) according to claim 2, characterized in that the criteria selected are the standard deviation of the intensities of the perpendicular image, the standard deviation of the intensities of the parallel image, and the ratio between the standard deviation of the intensities of the parallel image and the standard deviation of the intensities of the perpendicular image. 4) Decision device (100) according to claim 2, characterized in that the criteria used are the standard deviation of the intensities of the perpendicular image on which a high-pass type frequency filter is applied beforehand, the difference -type of the intensities of the parallel image on which a high-pass type frequency filter is previously applied, and the ratio between the standard deviation of the intensities of the parallel image on which a pass-type frequency filter is applied beforehand high and the standard deviation of the intensities of the perpendicular image on which a high pass type frequency filter is applied beforehand.
5) Decision device (100) according to claim 1, characterized in that the representative criteria of scab are characteristic dimensions of the scab observed and calculated by an autocorrelation calculation.
6) Decision device (100) according to claim 1 or 5, characterized in that the criteria representative of scab is the standard deviation of the power spectral density.
7) Decision device (100) according to one of claims 1 to 6 characterized in that the comparison means comprise the use of a learning mechanism of the ACP type, SVM or neuron network driven on bases of real tissue and fraud.
8) A decision method (300) provided for deciding whether a surface (50) is a living tissue or not and implemented by a decision device (100) according to one of claims 1 to 4, the decision process ( 300) comprising:
 a lighting step (302) during which the lighting means (102) illuminate the surface (50) to be analyzed,
a capture step (304) in which the perpendicular capture means (106) captures a perpendicular image of the surface (50) and illuminated and / or a capture step (306) during which the parallel capture means (108) captures a parallel image of the illuminated surface (50),
 a calculation step (306) during which the computing means (112) calculate the value of each of said criteria selected from each captured image,
 a comparison step (308) during which the comparison means (1 14) compares each value thus calculated with the interval of the corresponding reference values and delivers information representative of this comparison,
 - a decision step (310) in which the decision means (116) decide whether the surface (50) is a living tissue or not according to the information provided by the comparison means (114).
PCT/EP2014/062277 2013-06-12 2014-06-12 Decision device for deciding whether a surface is living tissue or not WO2014198847A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
FR1355428A FR3007170B1 (en) 2013-06-12 2013-06-12 Decision device provided for deciding if a face is true or false
FR1355428 2013-06-12

Publications (1)

Publication Number Publication Date
WO2014198847A1 true WO2014198847A1 (en) 2014-12-18

Family

ID=49274790

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2014/062277 WO2014198847A1 (en) 2013-06-12 2014-06-12 Decision device for deciding whether a surface is living tissue or not

Country Status (2)

Country Link
FR (1) FR3007170B1 (en)
WO (1) WO2014198847A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3139307A1 (en) * 2015-09-07 2017-03-08 Safran Identity & Security Method for characterising a material by speckle analysis
WO2019038128A1 (en) * 2017-08-22 2019-02-28 Lumileds Holding B.V. Laser speckle analysis for biometric authentication

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080025579A1 (en) * 2006-07-31 2008-01-31 Lumidigm, Inc. Spatial-spectral fingerprint spoof detection
US20080219522A1 (en) * 2007-03-09 2008-09-11 Authentec, Inc. Finger sensor using polarized light and associated methods
WO2008108871A2 (en) * 2006-07-31 2008-09-12 Lumidigm, Inc. Spatial-spectral fingerprint spoof detection
US20110163163A1 (en) * 2004-06-01 2011-07-07 Lumidigm, Inc. Multispectral barcode imaging
US20120163783A1 (en) * 2010-12-22 2012-06-28 Michael Braithwaite System and method for illuminating and imaging the iris of a person

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110163163A1 (en) * 2004-06-01 2011-07-07 Lumidigm, Inc. Multispectral barcode imaging
US20080025579A1 (en) * 2006-07-31 2008-01-31 Lumidigm, Inc. Spatial-spectral fingerprint spoof detection
WO2008108871A2 (en) * 2006-07-31 2008-09-12 Lumidigm, Inc. Spatial-spectral fingerprint spoof detection
US20080219522A1 (en) * 2007-03-09 2008-09-11 Authentec, Inc. Finger sensor using polarized light and associated methods
US20120163783A1 (en) * 2010-12-22 2012-06-28 Michael Braithwaite System and method for illuminating and imaging the iris of a person

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
FROM WIKIPEDIA: "Coherence length", 17 October 2014 (2014-10-17), XP055147486, Retrieved from the Internet <URL:http://en.wikipedia.org/wiki/Coherence_length> [retrieved on 20141017] *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3139307A1 (en) * 2015-09-07 2017-03-08 Safran Identity & Security Method for characterising a material by speckle analysis
FR3040815A1 (en) * 2015-09-07 2017-03-10 Morpho Method of characterizing a material by tavelure analysis
US9830520B2 (en) 2015-09-07 2017-11-28 Safran Identity & Security Method for characterizing material by analysis of speckles
WO2019038128A1 (en) * 2017-08-22 2019-02-28 Lumileds Holding B.V. Laser speckle analysis for biometric authentication

Also Published As

Publication number Publication date
FR3007170B1 (en) 2016-09-02
FR3007170A1 (en) 2014-12-19

Similar Documents

Publication Publication Date Title
Bowyer et al. Image understanding for iris biometrics: A survey
Zhang et al. An online system of multispectral palmprint verification
US8180120B2 (en) Finger sensor using polarized light and associated methods
US7347365B2 (en) Combined total-internal-reflectance and tissue imaging systems and methods
DE60211678T2 (en) Method and apparatus for biometric identification or biometric detection of individuals using optical spectroscopy
Kono et al. Near-infrared finger vein patterns for personal identification
US7801335B2 (en) Apparatus and methods for detecting the presence of a human eye
KR101356358B1 (en) Computer-implemented method and apparatus for biometric authentication based on images of an eye
JP4844939B2 (en) Spectroscopic methods and systems for multifactor biometric authentication
Marasco et al. A survey on antispoofing schemes for fingerprint recognition systems
JP4459137B2 (en) Image processing apparatus and method
EP1732028B1 (en) System and method for detecting an eye
US20030223621A1 (en) Methods and systems for biometric identification of individuals using linear optical spectroscopy
Lee et al. Fake iris detection by using purkinje image
US20090310827A1 (en) Method and device for recognition of natural skin
CN100419779C (en) Facial liveness assessment system
JP5805131B2 (en) Method and apparatus for providing strobe image capture
EP2981929B1 (en) Fingerprint pore analysis for liveness detection
KR101349892B1 (en) Multibiometric multispectral imager
EP1988489A1 (en) Biometric sensor apparatus and method
CN102339382B (en) Multispectral imaging biometrics
US20120213418A1 (en) Multimodal ocular biometric system and methods
US8781181B2 (en) Contactless multispectral biometric capture
Sinha Qualitative representations for recognition
US20090161925A1 (en) Method for acquiring the shape of the iris of an eye

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14734754

Country of ref document: EP

Kind code of ref document: A1

DPE1 Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101)
NENP Non-entry into the national phase in:

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14734754

Country of ref document: EP

Kind code of ref document: A1