CN100447804C - Methods and systems for estimation of personal characteristics from biometric measurements - Google Patents

Abstract

Methods and apparatus are provided for estimating a personal characteristic of an individual. A biometric data measurement is collected from the individual. The personal characteristic is determined by applying an algorithmic relationship between biometric data measurements and values of the personal characteristic derived from application of a multivariate algorithm to previous measurements.

Description

Come the method and system of estimation of personal characteristics according to biometric measurement

The application is the U.S. Provisional Patent Application No.60/529 that is entitled as " DEMOGRAPHIC INFORMATION ESTIMATION FROMDERMATOGLYPHIC AND OTHER BIOMETRIC PATTERNS " that was submitted by Robert K.Rowe on Dec 11st, 2003,299 non-provisional application, and require its rights and interests, in the lump as a reference at this with its content.

Technical field

The U.S. Patent application No.10/818 that is entitled as " MULTISPECTRAL BIOMETRIC SENSOR " that the application relates on April 5th, 2004 and submitted by Robert K.Rowe etc., 698, it is the non-provisional application of following each application: the U.S. Provisional Patent Application No.60/460 that is entitled as " NONTINVASIVE ALCOHOL MONITOR " that on April 4th, 2003 submitted, 247; The U.S. Provisional Patent Application No.60/483 that is entitled as " HYPERSPECTRAL FINGERPRINT READER " that on June 27th, 2003 was submitted by Robert K.Rowe etc., 281; The U.S. Provisional Patent Application No.60/504 that is entitled as " HYPERSPECTRALFINGERPRINTING " that on September 18th, 2003 submitted, 594; And on March 10th, 2004 the U.S. Provisional Patent Application No.60/552 that is entitled as " OPTICAL SKIN SENSOR FOR BIOMETRICS " that submits, 662.The U.S. Patent application No.09/874 that is entitled as " APPARATUS AND METHOD OFBIOMETRIC DETERMINATION USING SPECIALIZED OPTICALSPECTROSCOPY SYSTEM " that the application also relates to June 5 calendar year 2001 and submitted by Robert K.Rowe etc., 740, and the U.S. Patent application No.10/640 that is entitled as " ELECTRO-OPTICAL SENSOR " that relates on August 12nd, 2003 and submit by Robert K.Rowe etc., 503 (" ' 503 applications ").' 503 applications are non-provisional applications of following application: the United States Patent (USP) provisional application No.60/403 that is entitled as " BIOMETRIC ENROLLMENTSYSTEMS AND METHODS " that on August 13rd, 2002 was submitted by Robert K.Rowe etc., 453; The United States Patent (USP) provisional application No.60/403 that is entitled as " BIOMETRICCALIBRATION AND DATA ACQUISITION SYSTEMS ANDMETHODS " that on August 13rd, 2002 was submitted by Robert K.Rowe etc., 452; The United States Patent (USP) provisional application No.60/403 that is entitled as " BIOMETRIC SENSORS ONPORTABLE ELECTRONIC DEVICES " that on August 12nd, 2002 was submitted by Robert K.Rowe etc., 593; The United States Patent (USP) provisional application No.60/403 that is entitled as " ULTRAHIGH-SECURITY IDENTIFICATION SYSTEMS AND METHODS " that on August 13rd, 2002 was submitted by Robert K.Rowe etc., 461; The United States Patent (USP) provisional application No.60/403 that is entitled as " MULTIFUNCTION BIOMETRIC DEVICES " that on August 13rd, 2002 was submitted by Robert K.Rowe etc., 449; And on April 4th, 2003 the United States Patent (USP) provisional application No.60/460 that is entitled as " NONINVASIVE ALCOHOL MONITOR " that submits by Robert K.Rowe etc., 247.The whole of above-mentioned each application disclosed in the lump at this as a reference.

Background technology

The application relates generally to biostatistics.More specifically, the application for example relates to by estimating anthropological measuring, demography and/or according to the physiological parameter of fingerprint or other skin line (dermatoglyphic) image, estimating the personal characteristics's of individuality method and system according to biometric measurement.

The classic method of biometrical identification systems uses biometric measurement that individual unique identification is provided.For example, can collect for example biometrics of fingerprint in the scene of a crime by the law enfrocement official, and compare, so that identification suspect with fingerprint database.Use this information to search for the suspect who is discerned by law enforcement agency then, as a part of scouting, scouting can also comprise the motivation of determining scene of a crime and each suspect, so that set up the evidence-based of prosecution.Perhaps, other biometrical identification systems is devoted to confirm personnel identity of attempting access security mechanism etc.Except fingerprint is measured, exist a large amount of other to can be used for the biometric characteristic of these purposes, comprise facial and hand dimensional measurement, iris and retina scanning etc.

The restriction of the integrality of the database that the effect of these biometric identification methods is compared.For example, under the law enforcement sight, perhaps the fingerprint of collecting in the scene of a crime does not have homologue in any addressable fingerprint database, because the individual of crime is not gathered fingerprint as yet.Perhaps, current system has write down fingerprint characteristic for the evidence purpose of discerning the suspect, but when the coupling that do not exist with existing database, it is useless for identification suspect usually.

Similar problem appears under multiple other situation.For example, some or everyone that screen that requirement occurs at intersection of the immigrant of some countries now fingerprint of measuring them.If fingerprint of measuring and known criminal or terrorist's record are complementary, then screening can be taked suitable behavior.Similarly, this process is subject to the inevitable incomplete fact of existing fingerprint database.There is following possibility in this: personnel utilize a large amount of fingerprints of collection not have the fact of database homologue, show certain deceptive practices of different fingerprints by using reparation or miscellaneous equipment.For example, perhaps allowing known fingerprint is that 30 years old white man man of one of wanted criminal enters, because he has shown 65 years old soul sister's fingerprint with cheating, only because database this woman's fingerprint recording not.Even the screening right authority has valuable biometric information, it also can not effectively use this information.

In some cases, biometric measurement system can produce commercial interested or interested other estimation of individual subscriber.For example, provide biometrics can be used to cosmetics industry about the information of skin.Similarly, the biometric sensor of estimating the physiological parameter of blood perfusion for example or water cut is useful for the consumer in motion.

Therefore, technical needs can adapt to the improved biometrics analysis of these situations.

Summary of the invention

Therefore, embodiments of the invention provide the method and apparatus of estimating individual personal characteristics according to biometric measurement.In first group of embodiment, provide a kind of method that is used to estimate individual continuous personal characteristics.Measure from individual collection of biological statistics.By measure and use between the personal characteristics value continuously the algorithm relation at biometric data, determine individual continuous personal characteristics, wherein the personal characteristics value obtains from the personal characteristics value of a plurality of biometric datas measurements and corresponding collection is used algorithm continuously.

In more such embodiment, the biometric data measurement comprises the measurement of skin line.Skin part that can be by utilizing the rayed individuality and receive light from the skin part scattering makes and can obtain multispectral image from the light that receives, and comes the collection of biological statistics to measure.The example of personal characteristics comprises age, body weight, body-mass index, ethnic group, race and job classification continuously.In one embodiment, collect a plurality of biometric measurements and corresponding personal characteristics value, and the application algorithm obtains the algorithm relation.In another embodiment, by between the biometric data measurement and the second personal characteristics value, using second algorithm relation, determine second personal characteristics, wherein the second personal characteristics value obtains from the second personal characteristics value of a plurality of biometric datas measurements and corresponding collection is used second algorithm.Described algorithm can comprise polynary (multivariate) algorithm.

In second group of embodiment, provide a kind of method that is used to estimate individual personal characteristics.Utilize light to shine individual skin part.Reception is from the light of skin part scattering.From the light that receives, obtain multispectral image.By between multispectral image and personal characteristics value, using the algorithm relation, determine individual personal characteristics, wherein the personal characteristics value obtains from the personal characteristics value of a plurality of multispectral images and corresponding collection is used algorithm.

In more such embodiment, personal characteristics is the binary personal characteristics, and in other embodiments, personal characteristics is continuous personal characteristics.Can shine skin part by producing light in a plurality of discrete wave strong points as a plurality of quasi-monchromatic wave bundles, then with the optical registration skin part that produces.Alternatively, can shine skin part by producing broad band light beam and aiming at skin part a plurality of discrete wave strong point filtering broadband beams, general filter width wave beam.In some instances, can utilize the rayed skin part with first polarization, the light of reception has in fact second polarization that intersects with first polarization.In one embodiment, collect multispectral image and corresponding personal characteristics value, and the application algorithm obtains the algorithm relation.In another embodiment, by between the multispectral image and the second personal characteristics value, using second algorithm relation, determine second personal characteristics, wherein the second personal characteristics value obtains from the second personal characteristics value of a plurality of multispectral images and corresponding collection is used second algorithm.Algorithm can comprise polynary algorithm.

In the 3rd group of embodiment, provide a kind of method that is used to estimate individual personal characteristics.Measure from individual collection of biological statistics.By the algorithm relation between measurement of applying biological statistics and the personal characteristics value, determine personal characteristics, wherein the personal characteristics value obtains from the personal characteristics value of a plurality of biometric datas measurements and corresponding collection is used polynary algorithm.

In some instances, the biometric data measurement comprises the measurement of skin line.In one embodiment, the skin part by utilizing the rayed individuality and receive light from the skin part scattering makes and can obtain multispectral image from the light that receives, and comes the collection of biological statistics to measure.

Can utilize have illumination subsystem, detection subsystem and carry out certain methods of the present invention with the sensing system of the computing unit of detection subsystem interfaces.Place illumination subsystem and provide light to give individual skin part.Place detection subsystem to receive light from skin part.Computing unit has the instruction of implementing the invention described above method.

Description of drawings

With reference to the other parts and the accompanying drawing of instructions, can further understand essence of the present invention and advantage, all in the accompanying drawing, similar reference marker is used to indicate similar assembly.In some instances, reference marker comprises the numerical portion with Latin alphabet suffix, and only the quoting of numerical portion for reference marker is intended to unify to refer to have this numerical portion but all reference markers with different Latin alphabet suffix.

Fig. 1 shows the process flow diagram of the method that is used to estimate individual personal characteristics in embodiments of the present invention;

Fig. 2 shows the process flow diagram of the method that is used for the collection of biological statistical measurement that can use with the method for Fig. 1;

Fig. 3 provides the front view of used multispectral biometric sensor in the embodiment of the invention;

Fig. 4 A provides the side view of the multispectral biometric sensor shown in the embodiment;

Fig. 4 B provides the side view of the multispectral biometric sensor shown in another embodiment;

Fig. 5 provides the schematically illustrating of computer system of the function that can be used to manage the personal characteristics estimating system among the embodiment;

Fig. 6 provides the front view of computer tomography spectrometer (" CTIS ") in one embodiment of the present of invention;

Fig. 7 provides the top view of the striking in the embodiment of the invention (swipe) sensor; And

Fig. 8 shows the multispectral data cube (datacube) that produces according to the embodiment of the invention.

Embodiment

1. general introduction

Embodiments of the invention use the correlativity between biometric measurement and the personal characteristics, make can use this individual biometric measurement in the personal characteristics of estimating unknown individual.As used herein, " personal characteristics " is meant individual anthropological measuring, demography and/or physiological characteristic particularly." anthropological measuring " parameter can comprise individual many aspects, for example height, body weight or body-mass index or " handedness (handedness) " (promptly distinguishing the fingerprint that obtains from the right hand or left-hand finger)." demography " parameter can comprise plurality of classes, for example blood perfusion degree etc. in the amount of cutaneous pigmentation and type, skin oleaginousness, moisture content of skin, the skin.In some instances, herein with reference to " binary personal characteristics attribute ", it is to adopt the only attribute of one of two possibility states; For example, the example of binary personal characteristics attribute comprises sex and handedness.Herein sometimes also with reference to " personal characteristics attribute continuously ", it is usually about being worth unrestricted and can getting the attribute of arbitrary value (may in the specific endpoints restriction); The example of anthropological measuring attribute comprises age and body weight continuously.Depend on application, some personal characteristics attributes can be the function of binary function or conitnuous forms.For example, can be the binary personal characteristics with ethnic attribute definition, wherein the individual can or cannot be divided into and belong to particular race.Yet more generally, for example Zhong Zu personal characteristics is taken as continuous personal characteristics and handles so that recognize and distribute classification in any case, all exist a plurality of classification may, and most of individuality presents plurality of classes to a certain degree.

According to embodiments of the invention, the biometrics pattern can be used to set up the special characteristic of the grouping that is studied.For example, can set up the genetic distance between the particular race grouping by the pattern that relatively in biometric measurement, finds.Can on phenotype (genotypic individual the expression) rank, not carry out this estimation on genotype (grouping feature) rank.For example the specific gene disease of Down syndrome also is associated with particular organisms statistics pattern.

In certain embodiments, the biometrics pattern of use comprises skin line pattern, can collect skin line pattern from the skin part that comprises finger, palm, toe and sole.The proper method that is used to obtain this pattern comprises: the optical imagery based on total internal reflection (" TIR ") phenomenon is caught; Direct optical imagery; Capacitive radio frequency (" RF ") and other semiconductor array capture device; Ultrasonic, pressure array etc.Term " fingerprint " is used to broadly to represent have the expression of any skin part of skin line feature.In addition, can carry out optical acquisition according to the mode of measuring a plurality of optical conditions at identical skin part place.Optical system can comprise the multispectral and/or ultraphotic spectrum capture device that uses a plurality of illumination wavelengths to throw light on.Optical system can also be measured under two or more polarization conditions.The system that is collected in the image that obtains under a plurality of optical conditions of different wave length for example and/or different polarization condition is the example of " multispectral system ".A kind of detailed description of example of multispectral system is provided below, and this multispectral system can correspondingly be used in the embodiment of the invention, but because in alternative, can use other technology, so this description is not intended to be restriction.Advantageously, the collection of multispectral data is a robust for imperfect skin quality, for example usually with the elderly, carry out important blue-collar people or for example hairdresser or nurse's skin be exposed to drying that the people of chemicals is associated, lack elasticity and/or wan and sallow feature.In addition, except the data that alternative fingerprint sensing technology is provided, advantageously, this multispectral data can comprise carries out the additional information that personal characteristics estimates at usefulness.

The view of Fig. 1 provides total general introduction of method of the present invention.This method starts from setting up correlativity at first between personal characteristics and biometric measurement.Shown insquare frame 104, this can come the collection of biological statistics to realize by a plurality of people being carried out biometric measurement.Collect for example personal characteristic information of age, sex, handedness, body weight etc. atsquare frame 108 places.Preferably, personnel's grouping of selecting at this collection shows the personal characteristics attribute of wider range, especially for continuous personal characteristics attribute, and for example comprises the extremum of age on a large scale, very high and low-down body weight etc.Usually the initial input that utilizes balance more be used for relevant quite a large amount of measurement, improve the reliability of this method.In certain embodiments, the personnel's number in the grouping is greater than 1000.

Atsquare frame 112 places,, the data of collecting are carried out multivariate analysis, so that biometric data is associated with anthropological measuring information for personnel's grouping.This analysis can comprise the application of various preconditioning techniques, with the number that reduces feature, improve picture quality and/or realize the feature of other hope.For example, can use the intrinsic analysis, to gather and describe each biometric measurement by analyzing corresponding mark with the intrinsic of selected number to biometric data.The attributes extraction method can comprise intrinsic analysis, linear discriminant analysis and view data is decomposed into a plurality of basis functions of setting up in advance that comprise sine and cosine (frequency analysis), small echo, Gabor wave filter, radial basis function, and technical other method of having known.Comprise spectral measurement in particular cases at biometric measurement, preconditioning technique can extract for example feature of piece spectral signature, corresponding ratings etc.In some cases, it is favourable using for the constant feature of the specific effect that takes place during data aggregation.For example, under given conditions, frequency analysis can produce the mobile constant feature for the object of wanting imaging.In some instances, use data rather than the pretreated data directly measured advantageously to make the application of polynary algorithm can pick out the meticulous correlativity and the spectrum relation of perhaps bluring in the pattern by the pre-service application.

For example, can be applied to attribute and/or raw data comprises with the example of the multiplex method of the correlativity of personal characteristics determining: linearity or quadratic discriminatory analysis, part least square analysis, principal component recurrence etc.In some instances, can realize the application of multiplex method by the training evaluating system of for example expert system, neural network etc.The application of multiplex method makes the biometric data of measurement be associated with the anthropometric data of collection, so that the definition correlationship.For example, the mark set that the intrinsic analysis according to the biometric data of measuring is produced is associated with personal characteristics, thereby provides the algorithm that both are associated to concern.Can carry out this processing to binary anthropological measuring attribute and continuous anthropological measuring attribute.Under the situation of the binary anthropological measuring attribute of for example sex or handedness, can use quantitative regression technique or sorting technique to set up correlativity.The proper technology that biometric measurement is associated with the non-quantitation personal characteristics of for example ethnic group comprises k nearest neighbor technology and other clustering technique.

Aftersquare frame 112 places have set up algorithm relation, can carry out feature to the biometric measurement in the correlativity set not and estimate.Under the situation of the continuous personal characteristics of for example age and body weight, estimate to comprise interpolation or extrapolation correlation model, although, then seldom need extrapolation method usually if initial biometric measurement set is a balance.Therefore, atsquare frame 116 places, usually by repeating and setting up the identical measurement of carrying out in the correlative relationship of measurement, from target individual collection of biological statistics at first.Atsquare frame 120 places, by the attribute that extracts is applied in the algorithm relation of setting up in thesquare frame 112, correspondingly estimation of personal feature from biometric data to the biometric data of collection and/or according to the same way as that is applied to the correlation data collection.

In some instances, can from the collection of biological statistics of single set, obtain a plurality of correlative relationships, utilize each correlative relationship to set up algorithms of different relation from the data measured to different personal characteristics.For example, have the range of age when initial personnel set comprises, during the personnel of different sexes, different ethnic group, different weight etc., can set up different algorithm relations about these different attributes.The existence of this correlativity makes it possible atsquare frame 120 places, sets up a plurality of personal characteristics of target individual according to the measurement of carrying out atsquare frame 116 places.

For from the skin part collection of biological statistics of individuality embodiment as multispectral data, the process flow diagram of Fig. 2 provide atsquare frame 104 places to a plurality of individualities or atsquare frame 116 places to how total general introduction of collection of biological statistics of target individual.Atsquare frame 204 places, the light with a plurality of wavelength is provided, and carries out polarization at this light ofsquare frame 108 places.Atsquare frame 212 places, directional lighting light is to shine individual skin part.At square frame 126 places, receive light, and this light is carried out polarization atsquare frame 220 places from the skin part scattering.The tissue thickness that may depend on skin part, scattered light can comprise reflection and/or transmitted light usually.The light that provides atsquare frame 204 places comprises under the situation of a plurality of wavelength, carries out spectral separation atsquare frame 224 places to receiving light, makes different wave length to be directed to photodetector atsquare frame 228 places.Then, the light of collecting at the photodetector place is used to obtain being used to the multispectral image of the relevant or estimation of personal characteristics as mentioned above.

Usually, if illumination light atsquare frame 208 places without polarization, the major part of detected light will comprise from the optical energy of skin surface or very shallow degree of depth reflection in skin.On the contrary, if use polarizer to come polarized light atsquare frame 208 places, and use second (quadrature) polarizer atsquare frame 220 places, then detected light is because the optics under the skin surface interacts basically.This cross polarization system can comprise it is orientated so that the axle of two polarizers separates two about 90 ° linear polarizers.Alternatively, polarizer can be a circular polarisers, wherein realizes quadrature by two circuit polarizers with opposite perception (being the right hand and left hand).Because even the effect of polarizer when using identical illumination wavelengths, also can be collected a plurality of different images by the polarization state that changes system.

In specific example, only in the cubical part of whole multispectral data, comprise the information of hope.For example, perhaps the estimation of equally distributed spectrum activity material only needs the measure spectrum feature that can extract from whole multispectral data cube.In this case, can simplify the total system design, to reduce or to remove the spatial component of collecting data by reducing the image pixel number limit of a pixel (or even reduce to).Therefore,, can recognize, the present invention includes the similar measurement that significantly reduces imaging degree (or even reducing to detector cell of existence) although in the context of multispectral imaging, described disclosed system and method.

Multispectral biometric sensor

Described to be used for a kind of structure of the multispectral sensor of collection of biological statistics with reference to the synoptic diagram of figure 3.Multispectral sensor 301 comprises:illumination subsystems 321 has one or more light sources 303; Anddetection subsystem 323, have imager 315.The figure shows the embodiment thatillumination subsystems 321 comprises a plurality ofillumination subsystems 321a and 321b, but the present invention is not limited to the number of this illumination ordetection subsystem 321 or 323.For example, can select the number ofillumination subsystems 321 easily, to realize the illumination of specific degrees, encapsulate requirement, and satisfy other structural limitations of multispectralbiometric sensor 301 in order to satisfy.As another example, can there be a plurality ofdetector subsystem 323 of arranging especially, in one or more subsystems, to incorporate the different optical effect of for example total internal reflection into according to different modes.303 by illumination optics 305, optical device 305 makes illumination form the form of wishing, for example form of floodlight (flood light), light, luminous point etc. to illumination light from the source.For convenience, illumination optics 305 is shown as and comprises lens, but it comprises the combination in any of one or more lens, one or more catoptron and/or other optical element usually.Illumination optics 305 can also comprise the scanning mechanism (not shown), scans illumination light in order to according to the rules one dimension or two-dimentional pattern.In different embodiment, light source 303 can comprise pointolite, line source, area source, perhaps can comprise a series of this light sources.In one embodiment, for example by placinglinear polarizer 307, by this polarizer, illumination light is provided as polarized light to light before thefinger 319 that arrives the personnel that will study or other skin part.

In some instances, light source 303 can be included in one or more quasi-monochromatic sources that light is provided on the narrow wavelength range.This quasi-monochromatic source can comprise for example device of light emitting diode, laser diode or quantum dot laser.Alternatively, light source 303 can comprise the wideband light source of incandescent lamp bulb for example or light-emitting section.Under the situation of wideband light source, illumination light can be by bandpass filter 309 so that the spectral width of illumination light narrow down.In one embodiment, bandpass filter 309 comprises one or more dissociated optical bandpass filter.In another embodiment, bandpass filter 309 comprises the continuously variable filter of circulation or linear moving (perhaps having circulation and linearly moving combination), in order to change the illumination light wavelength.In another embodiment, bandpass filter 309 comprises tunable optic filter elements, for example liquid crystal tunable filter, acousto-optic tunable filter, tunable Fabry-Perot wave filter or well known to a person skilled in the art other filter mechanism.

After the light of light source 303 passed through illumination optics 305 and passes through optical filter 309 and/orpolarizer 307 alternatively, this light was by pressingplate 317, andirradiation finger 319 or other skin part.Advantageously, can select sensor placement and assembly, so that be directed to the direct reflection minimized of the illumination of detection optics 313.In one embodiment,, make detected direct catoptrical amount minimize, reduce this direct reflection by relatively being orientatedillumination subsystems 321 and detection subsystem 323.For example, can place the optical axis ofillumination subsystems 321 anddetection subsystem 323 at a certain angle, so that the catoptron that is placed on thepressing plate 317 does not make the illumination light of obvious amount shine detection subsystem 323.In addition, can be to place the optical axis of illumination anddetection subsystem 321 and 323 with respect to pressingplate 317 certain angle ground, make the critical angle of two acceptable angles of subsystem less than system; But the perceived effect that causes owing to the TIR betweenpressing plate 317 and theskin part 319 has been avoided in this configuration.

The optional mechanism that is used to reduce the light of direct reflection is to use optical polarizator.As known for the skilled artisan, can advantageously adopt linear and circular polarisers, so that optical measurement is sensitive more for the particular skin degree of depth.In the embodiment shown in fig. 3, come illumination light is carried out polarization by linear polarizer 307.Thendetection subsystem 323 can also compriselinear polarizer 311, and its optical axis is set in fact andillumination polarizer 307 quadratures.In this manner, must experience repeatedly scattering events from the light of sample, so that change its polarization state significantly.When being scattered backdetection subsystem 323 when light transdermal surface and after scattering events repeatedly, this incident appears.In this manner, reduced the surface reflection at the interface betweenpressing plate 317 andskin part 319.

Detection subsystem 323 can merge detection optics, and described detection optics is included in lens, catoptron and/or other optical element of the image that is formed onclamp surface 317 near zones on the detecting device 315.Detection optics 313 can also comprise the scanning mechanism (not shown), in order to successively the part of platen region is passed to detecting device 315.In all cases,detection subsystem 323 is configured to for the transdermal surface and in the light sensitivity of passing through optical scattering before the skin outgoing in skin and/or hypodermis.In some cases, light source 303 can be the wideband light source that does not need spectral filter 309 to use as illumination subsystems.On the contrary, the color filter array that comprises the microarray of different band bandpass filter directly can be incorporated on the pattern matrix 315.The common color filter that appears on the multiple colour imaging chip is the Bayer wave filter, and as known to those skilled in the art, it has described the layout of redness, green and blue ribbon bandpass filter.

As mentioned above, it is favourable measuring the image that obtains under the different polarization condition.By with reference to two illumination subsystems 312a and 312b, check the example of the mode of so doing.In this embodiment, an illumination subsystems 312a incorporateslinear polarizer 307a with respect to the cross polarization condition that detectspolarizer 311 into.Thesecond illumination subsystems 321b has omitted linear polarizer 307b.In this configuration, can utilize theillumination subsystems 321a of polarization to collect first image, this image is illustrated in optical scattering and other effect underskin 319 surfaces in fact.Then, utilizeunpolarized illumination subsystems 321b to collect second image.Althoughpolarizer 311 is arranged indetection subsystem 323, the illumination light in this second image of polarization not, and the image section that produces ground derives from the very shallow scattering of surface reflection and light, and from the degree of depth scattering of pointing 319 light.In the estimation of personal characteristics, the combination of two images can be used to provide additional useful information.

Illumination subsystems 321 anddetection subsystem 323 can be configured to work in multiple optical condition and a plurality of wavelength.An embodiment uses emission to be in the light source 303 of the light of 400-1000nm scope basically; In this case, detectingdevice 315 can be based on silicon detector element or other detector material for this wavelength-sensitive known in those skilled in the art.In another embodiment, light source 303 can be in the wavelength place emitted radiation of the near infrared condition that comprises 1.0-2.5 μ m, in this case, detectingdevice 315 can comprise the element of being made up of other material of InGaAs, InSb, PbS, MCT and the light sensitivity for this wavelength known in those skilled in the art.

Fig. 4 A has illustrated the side view of the embodiment of the invention with synoptic diagram.For clear, this view and not shown detection subsystem, but clearly show that illumination subsystems 321.In the present embodiment,illumination subsystems 321 comprises two separate light source 403 and 405 that can have the different wave length feature.For example, light source 403 and 405 can be the quasi-monochromatic source that does not need optical filter of LED forexample.Light source 403a, 403b and 403c can provide the illumination of identical in fact first wavelength, andlight source 405a, 405b and 405c can provide the illumination of identical in fact second wavelength, and second wavelength is different with first wavelength.As shown in the figure, illumination optics among Fig. 4 A is configured to provide floodlighting, but in optional embodiment, it can be configured to provide the illumination of line, point or other pattern by merging cylinder optical device, focusing optics or other optical module known in those skilled in the art.As previously described,polarizer 307 can appear on all light sources 403,405, and can appear on the part of light source 403,405, perhaps can omit fully.

The typical case of system shown in Fig. 4 A measures sequence and comprises the image that activates first light source 403 and collect generation.After obtaining image, close first light source 403, and activate secondary light source 405, and collect the image that produces at different wavelength places.For the sensor that has more than an optical source wavelength, can repeat this illumination measurement sequence at employed all different wave lengths in the sensor.Can also recognize, can in the embodiment that determines the optical wavelength feature by the state of tunable optical filter, variable optical filter, removable dissociated optical wave filter etc., use identical in fact sequence.Alternatively, ifdetector array 315 comprises color filter array, then can open the light source 403,405 of different wave length simultaneously.Another the optional mechanism that is used to collect the image at different wave length place comprises and incorporates coding method into, thereby discern the light of each wavelength when at a plurality of wavelength of place's irradiation preset time.Then, collect the data of whole illumination sequence in such a way:, determine each wavelength response according to the coding that uses method as well known to those skilled in the art.Therefore, lighting engineering comprises round-robin method, division modulation, Hadamard coding etc.

The light illumination sequence of different measuring can be different.Can propose this variability in order to prevent Replay Attack (replay attack), in Replay Attack, write down effective signal set and playback afterwards, use so that biometric sensor lost efficacy.Can also extend the measurement variability between the sample in certain embodiments, wherein only use the subclass of available illumination wavelengths, compare with the respective subset of registration data intensive data then.

In fact light source 403 and 405 array need not be the plane shown in Fig. 4 A.For example, in other embodiments, optical fiber, fibrous bundle or fibre optic faceplate or wick (taper) can be delivered to light on the illuminated area from being positioned at the light source that makes things convenient for the position, and on illuminated area, light imaging once more (reimage) is to finger.Can be by as LED opening and closing drive current, control light source, alternatively, if use incandescent source, use the spatial light modulator of some forms of liquid crystal modulator for example or use microelectromechanical-systems (" MEMS ") technology to control aperture, reflective mirror or other this optical element, realize the quick switching of light.

For example the use of the optical element of optical fiber and fibrous bundle makes the structure can simplify multispectral biometric sensor.An embodiment has been shown in Fig. 4 B, and Fig. 4 B shows the electric scanning of the light source of the use of optical fiber and for example LED.Eachfiber 416a makes each LED that is positioned atillumination array 410 places link to each other with imaging surface, and the light thatother fiber 416b will reflectpasses imaging device 412 back, andimaging device 412 can comprise photodiode array or ccd array.Therefore the set ofwave filter 416a and 416b defines employedfibre bundle 414 in relaying light.

Enforcement in conjunction with the described method of Fig. 1 and 2 can be cooperated with computer system, described computer system with as link to each other or integrated in conjunction with the biometric sensor of one of described multispectral biometric sensor of Fig. 3-4B etc.Setting shown in Figure 5 comprises a plurality of assemblies that are applicable to bigger system; The mini-system of integrated portable set can use a spot of assembly.Fig. 5 generally shows how to implement each system element according to separation or more integratedmode.Computing equipment 500 is shown as and comprises the hardware element that is electrically connected and links to each other withbiometric sensor 517 through bus 526.In some instances,biometric sensor 517 comprises multispectral biometric sensor, but this not necessarily, and in other embodiments,biometric sensor 517 can be another kind of fingerprint sensor or other biometric sensor.Hardware element comprises theprocessing accelerator module 516 and thestorer 518 ofprocessor 502,input equipment 504,output device 506,memory device 508, computer-readablestorage media reader 510a,communication system 514, for example DSP or application specific processor.Computer-readablestorage media reader 510a also links to each other with computer-readable recording medium 510b, and that this combination is synthetically represented is long-range, local, fixing and/or dismountable memory device adds the storage medium that is used for comprising temporarily and/or more enduringly computer-readable information.Thatcommunication system 514 can comprise is wired, wireless, the interface of modulation and/or other kind connects, and make can with the external unit swap data.

Computing equipment 500 also comprises software element, and this software element is shown as and is currently located in theworking storage 520, comprisesoperating system 524 and for example is designed forother code 522 of the program of implementing the inventive method.It is obvious to the skilled person that according to specific (special) requirements, can use the variation of essence.For example, perhaps also use customization hardware and/or realize particular element with hardware, software (comprising for example portable software of applet) or the two.In addition, can adopt and being connected of other computing equipment of for example network input-output apparatus.

Fig. 6 schematically shows the front view of the another kind of structure that can be used to implement multispectral biometric sensor.In the present embodiment, multispectralbiometric sensor 601 comprises broadband illumination subsystem 623 and detection subsystem 625.For in conjunction with the described embodiment of Fig. 3, can there be a plurality of illumination subsystems 623 in certain embodiments, wherein Fig. 6 shows the specific embodiments with two illumination subsystems 623.The light source 603 that illumination subsystems 623 is comprised is broadband illumination light sources of incandescent lamp bulb or light-emitting section for example, perhaps can be to well known to a person skilled in the art other broadband illumination light source.The light of light source 603 is by illumination optical device 605 and linear polarizer 607, and alternatively can be by being used for that optical wavelength is limited in bandpass filter 609 in the particular range.Light is by pressing plate 117 and be directed to skin part 119.Part light is directed todetection subsystem 625 from skin 119 reflections, anddetection subsystem 625 comprises imagingoptic element 615 and 619, thelinear polarizer 611 of intersection and dispersing optics element 613.Dispersion element 613 can comprise that transmission-type or reflective one dimension or two-dimensional grating, prism or the light path that makes as well known to those skilled in the art are as the function of optical wavelength and other optical element that departs from.In the embodiment shown, first imagingoptic element 619 is used to collimate the light fromskin 619 reflections, makes it bycross spider polarizer 611 and dispersion element 613.The spectrum component of light with angular separation, and focuses on detectingdevice 617 by second imagingoptic element 615 respectively by dispersion element 613.As discussing in conjunction with Fig. 3, thepolarizer 607 and 611 that illumination anddetection subsystem 623 and 625 comprise respectively is used to reduce the direct catoptrical detection at detectingdevice 617 places.

Therefore, the multispectral image that produces according to the light that receives at the detecting device place is according to computed tomography imaging spectrometer (" CTIS ") mode " coding " image.In the image that produces, there are wavelength and spatial information simultaneously.Can obtain each spectral line by mathematical inversion or " reconstruct " of coded image.

Above is examples of " zone (area) " sensor configuration in conjunction with Fig. 3 and 6 described embodiment.Except this area transducer configuration, in certain embodiments, the multispectral imaging sensor can be configured to " striking " sensor.Fig. 7 schematically shows the top view of an example of swipe sensor.In the figure, the field ofillumination 703 ofsensor 701 and surveyedarea 705 conllinear in fact.In some embodiment ofswipe sensor 701, can exist more than single field of illumination.For example, can there be a plurality of field of illuminations on any side that is arranged in surveyed area 705.In certain embodiments, field ofillumination 703 can be partially or fully overlapping with surveyed area 705.By shown in Fig. 7 arrow, making finger or other body part striking optics effective coverage, utilizesensor 701 to collect the multispectral image data.The corresponding linear sensor can be stationary system or roller (roller) system, can also comprise scrambler in order to record position information and help to sew up complete two dimensional image, as known to those skilled in the art according to the image block sequence that produces.When using roll system, finger tip or other skin part can be rolled on the roller transparent for used optical wavelength.Successively from the discrete portions reception light of skin part, set up multispectral image then according to the light that partly receives from difference.

The polarizer that some embodiment comprised can also be used to create or further strengthen surface characteristics.For example, if edge and the direction polarization illumination light of the pressing plate parallel (" P ") of sampling and the polarizer that detection subsystem has been incorporated vertical orientated (" S ") into, then reflected light is blocked the nearly right extinction ratio of polarizer.Yet the light that is directed to fingerprint ridge (ridge) some place is by scattering optically, and this makes the polarization randomization effectively.This makes to be observed by the S polarized imaging system and absorbs and the part of radiative 50% magnitude again.

Above is illustrative in conjunction with the described system of specific embodiments, and is not intended to be restriction.There are multiple change and the alternatives that are in the object of the invention scope for above-mentioned example embodiment.In a plurality of examples, do not influencing in fact under the situation of function aspects of the present invention, can change the layout or the order of optical element.For example, in the embodiment that uses broadband illumination light source and one or more optical filters, wave filter can be arranged in any a plurality of somes place of illumination and detection subsystem.In addition,, it is evident that, can under discontiguous situation, carry out identical measurement in fact although accompanying drawing shows measured finger or other skin part contacts with pressing plate.In this case, the optical system configurations that can throw light on being used to and detect is used in a distance irradiation and makes the skin part imaging.The U.S. Provisional Patent Application No.60/552 that is entitled as " OPTICAL SKIN SENSOR FOR BIOMETRICS " that submits on March 10th, 2004 provides some examples of this system in 662.

The foregoing description is created in the image collection of the skin part under different wave length and/or the polarization conditions, perhaps produces the data of using reconfiguration technique can produce image collection, and is the same with the particular case of CTIS or coded illumination subsystem.In order to demonstrate, gather in conjunction with this spectrum picture and to carry out following discussion, do not need to produce these image collections although in the embodiment that does not directly produce these image collections, handle for biometrics subsequently.Fig. 8 shows the multispectral image set of demonstration, this sets definitionmultispectral data cube 801.

A kind of mode of decomposition data cube 801 is to make it to be decomposed into and to measure to be used to throw light on the wavelength of sample and/or each corresponding image of polarization conditions in the processing.In the accompanying drawings, show five separate pictures 803,805,807,809 and 811, with five discrete illumination wavelengths and/or lighting condition (for example, at the lighting point light source at position X, Y place; Exist/do not have polarization of illumination) corresponding.In the embodiment that uses visible light, image can be for example corresponding at the image that light produced at 450nm, 500nm, 550nm, 600nm and 650nm place with use.The optical effect of each image representative and the synergistic special wavelength of skin, and under the situation of the embodiment that measuring element skin contacts with pressing plate, the representative of each image interacts with skin and the optical effect of the light of the special wavelength by skin-pressing plate interface.Because by the optical properties and the skin components of the skin of wavelength shift, each multispectral image 803,805,807,809 and 811 differs from one another usually.For example, the wavelength that is shorter than approximate 600nm is by the blood strong absorption, and absorption peak is at approximate 540nm and 576nm.Image at these wavelength places shows the blood attribute consumingly, comprises when finger presses is on sensor surface turning white of finger, and the mottled pattern that causes owing to depths blood vessel more of part.The light source that wavelength is longer than approximate 600nm is insensitive for blood, and level and smooth in fact more and even.

Therefore data cube can be expressed as R (XS, YS, XI, YI λ), and has described when at light source point XS, when YS place is shone at each picture point XI, the amount of the scattered reflection light of the wavelength X that the YI place views.Can by in suitable light source point position to a response summation, sum up different illumination arrangement (general type, line style, etc.).Traditional non-TIR fingerprint image F (XI YI) can be described as the multispectral data cube of setted wavelength λ o imprecisely, and be sued for peace on all light source positions:

$F(X_I,Y_I)=\underset{Y_S}{\mathrm{\Σ}}\underset{X_S}{\mathrm{\Σ}}R(X_S,Y_S,X_I,Y_I,{\mathrm{\λ}}_0)$

On the contrary, spectrum sets of biometric data S (λ) makes poor between the measurement light intensity of setted wavelength λ and illumination and the detection position

Be associated:

$S(\stackrel{\→}{D},\mathrm{\λ})=R(X_I-X_S,Y_I-Y_S,\mathrm{\λ})$

Therefore, multispectral data cube R is associated with conventional fingerprint image and spectrum sets of biometric data.Multispectral data cube R is any one a superset (superset) of other two data sets, and is included in correlativity or the out of Memory that may lose in two independent forms.

Because the dioptrics index of laminate material and skin is also not obvious different on used wavelength coverage usually, and optical interface does not have the variation of essence during measuring intervals of TIME, all is identical at all wavelengths place in fact so the optics at skin-pressing plate interface interacts.Light from skin to pressing plate and the light from the pressing plate to skin are subjected to the influence of Fresnel reflections at optical interface.Therefore, compare with the light by air gap not, the light that passes air gap in receiver media is not strong.This phenomenon has formed the part of the image information that comprises in the multispectral data cube.

The multispectral image data cube comprises the spatial spectral information from multiple source.As just example, for finger tip that pressing plate contacts on situation about measuring, the data cube of generation comprises the effect that causes owing to following: (i) optical interface between finger tip and the pressing plate, with the info class that comprises in the non-TIR fingerprint of tradition seemingly; (ii) Zu Zhi overall spectrum feature, different people is different; (iii) near the blood vessel of skin surface, and especially be located immediately at kapillary under the friction-ridge of forming outside fingerprint pattern; And (iv) according to being distributed in blood vessel and other spectral effective structure of organizing the depths with the similar mode of vein imaging.Similarly, embodiments of the invention provide the multiple source that is used in measured finger tip or other skin part to extract the mechanism of biometric data, thereby multiple-factor biometrics Application in Sensing is provided.

Because the wavelength of skin and hypodermic complexity relies on attribute, have clear qualification and spectral signature clearly with the corresponding spectral value set in given picture position.These spectral signatures can be used to by the pixel ground multispectral image of classifying.Can carry out this assessment by from qualified image collection, producing typical organization's spectral quality.For example, can write down multispectral data shown in Figure 8 with N * 5 matrixes, wherein N comprises from living tissue rather than from the image pixel number of the data in surrounding air zone.Intrinsic analysis or other factor analysis performed for this group matrix produce the representative spectral signature that these organize pixel.Use for example Mahalanobies distance and residual the measuring of spectrum (metrics) then, concentrate the spectrum of pixel to compare subsequent data with the spectral signature of this previous foundation.If more than few images pixel has and the inconsistent spectral quality of living tissue, think that then the sample right and wrong are true and underproof, therefore provide to be used for incorporating anti-deception measures into mechanism according to the sensor of determining of sample activity.

Similarly, be among the embodiment of finger tip at sample, according to spectral quality, the multispectral image pixel is classified as " ridge (ridge) ", " paddy (velley) " or " other ".Can use discriminant analysis method to carry out this classification, for example linear discriminant analysis of described method, quadratic discriminatory analysis, principal component analysis, neural network and well known to a person skilled in the art other method.Because ridge and grain weevil element are adjacency on typical finger tip, so in some instances, the multispectral data of the part around the interested image pixel is used to the classified image pixel.In this manner, extract traditional fingerprint image, be used for further handling and the biometrics estimation from sensor.The classification of " other " can be indicated the image pixel that has with the desired different spectral quality of authentic specimen.The threshold value of the number of total picture element that is classified as " other " in the image can be set.If exceed this threshold value, determine that then sample is fict, and make suitable indication, and take action.

Can use whole data cube or its particular portion to assign to carry out the estimation of personal characteristics.For example, can use the lower spatial frequency information that suitable spatial filter is isolated dark spectral effective structure in the common representative tissue.Can use above-mentioned similar spatial frequency to separate and/or the pixel sorting technique data that take the fingerprint.Can in the manner described above spectral information be separated with the live part of image.Then, cubical this three part of deal with data, and use and well known to a person skilled in the art method, it is compared with related data.According to the intensity of relevant these features in the biometric data,, make judgement about the estimation of corresponding human bulk measurement feature.

3. example

There is the different useful application that to implement the inventive method and system in a large number.Particular example discussed herein does not become restriction only as illustrative.Each example is used in conjunction with the relation of the algorithm between described biometric data of Fig. 1 and the personal characteristics, and described algorithm relation is to produce according to biometric data of collecting and the corresponding personal characteristic information that shows various age, body weight, race and job classification (office work, work on the spot, machine work etc.) and have personnel of different sexes and handedness.

The relation that produces can be used to exectorial background, for example under the situation of the indistinct finger mark of retrieval as the part of crime detection.So even owing to do not store fingerprint can not the Direct Recognition suspect time in any enforcement database, it is relevant to use said method that fingerprint and suspect's sex, race, handedness and approximate age, body weight and job classification are carried out, and analyzes fingerprint.The number of the personal characteristics that can define is big more, then can make the law enfrocement official reduce the size of suspect storehouse (pool) widely.In addition, this information can combine with other scouting information that the law enfrocement official collects, this information relate to as the relevant effort of suspect's particular individual.The additional information that obtains from the correlativity with personal characteristics can and under the situation of other situation of eliminating, be arrested and/or convict so that scout and make great efforts more effectively.

The algorithm relation can also be used to migrate background.For example, foreign nationals can be applied for the entry visa of a country, the file of these personnel's that witness age, nationality, sex, job classification etc.Authenticate this document although can make efforts, the correlativity of the anthropological measuring attribute that can the application of the invention embodiment be provided is implemented this authentication.For example, a people can produce one's proof, and he is 30 years old male sex physical labourer's a file.These personnel's multispectral biometric measurement can be used to obtain the estimation of human body attribute, shows that these personnel are actual to be 38 years old and to have the feature consistent with the pencil pusher.This difference then can be used to before granting visa these personnel to be carried out further bar and look into.

Entering one when national, for example a people uses deception mechanism to hide the situation of its identity, the use similar procedure with attempting duplicity.This can be for example when immigration right authority measure fingerprint so that the situation of comparing with known terrorist's database.If represent of me to have and the inconsistent anthropological measuring attribute of its physical property, then can indicate existence deception mechanism, and need further to investigate.For example, have the correlativity with 70 years old women of different ethnic groups, then before permission enters, need these personnel are further investigated if look like 30 years old man's of specific ethnic group a people.

The algorithm relation can also be used to the particular business background, for example by using it auxiliary as the marketing for targeted advertisements.In multiple example, the particular demographic grouping of known identification age, sex and/or ethnic group has different reflections to different types of adverstising.Individual biometric is measured and is made and can come classified advertisement according to the correlativity of anthropological measuring attribute, thereby uses the most effective advertisement of expection.

After reading the disclosure, other application is conspicuous to those skilled in the art.

Therefore, for described several embodiment, for those of ordinary skills, only otherwise depart from restriction of the present invention, can make change, structural modification and equivalence.Therefore, above-mentioned explanation is not a limited field of the present invention, and described limited field is limited by claims.

Claims (33)

1. method of estimating individual continuous personal characteristics, described method comprises:

Measure from individual collection of biological statistics; And

By measure and use between the personal characteristics value continuously the algorithm relation at biometric data, determine individual continuous personal characteristics, described continuous personal characteristics value obtains from the personal characteristics value of a plurality of biometric datas measurements and corresponding collection is used algorithm.

2. method according to claim 1, wherein, one or more biometric data measurements comprise the measurement of skin line.

3. method according to claim 1, wherein, the measurement of collection of biological statistics comprises:

Utilize the skin part of rayed individuality;

Reception is from the light of skin part scattering; And

From the light that receives, obtain multispectral image.

4. method according to claim 1, wherein, continuously personal characteristics is from by selecting the following group that constitutes: age, body weight, body-mass index, ethnic group, race and job classification.

5. method according to claim 1 also comprises:

Collecting a plurality of biometric datas measures and corresponding personal characteristics value; And

A plurality of biometric datas measurements and corresponding personal characteristics value application algorithm are obtained the algorithm relation.

6. method according to claim 1, also comprise: by between the biometric data measurement and the second personal characteristics value, using second algorithm relation, determine the second individual personal characteristics, wherein the second personal characteristics value obtains from the second personal characteristics value of a plurality of biometric datas measurements and corresponding collection is used second algorithm.

7. method according to claim 1, wherein, described algorithm comprises polynary algorithm.

8. according to claim 1, one of 2 or 4 to 7 described methods, wherein, described biometric data measurement is that the spatial spectrum biometric data is measured.

9. method according to claim 3, wherein, described biometric data measurement is that the spatial spectrum biometric data is measured, described multispectral image is the spatial spectrum multispectral image.

10. method of estimating individual personal characteristics, described method comprises:

Utilize the skin part of rayed individuality;

Reception is from the light of skin part scattering;

From the light that receives, obtain multispectral image; And

By between multispectral image and personal characteristics value, using the algorithm relation, determine individual personal characteristics, wherein the personal characteristics value obtains from the personal characteristics value of a plurality of multispectral images and corresponding collection is used algorithm.

11. method according to claim 10, wherein, personal characteristics is the binary personal characteristics.

12. method according to claim 10, wherein, personal characteristics is continuous personal characteristics.

13. method according to claim 10, wherein, the irradiation skin part comprises:

Produce light in a plurality of discrete wave strong points, as a plurality of quasimonochromatic beam of light; And

Make the rayed skin part of generation.

14. method according to claim 10, wherein, the irradiation skin part comprises:

Produce broad band light beam;

Broad band light beam is carried out filtering, so that provide light in a plurality of discrete wave strong points; And

Make the broad band light beam irradiation skin part of filtering.

15. method according to claim 10, wherein:

The irradiation skin part comprises and utilizes first polarization to make light polarization; And

Receive light and comprise the light polarization of utilizing second polarization to make reception, described second polarization and first polarization cross.

16. method according to claim 10 also comprises:

Collect a plurality of multispectral images and corresponding personal characteristics value; And

A plurality of multispectral images and corresponding personal characteristics value are used algorithm, obtain the algorithm relation.

17. method according to claim 10, also comprise: by between the multispectral image and the second personal characteristics value, using second algorithm relation, determine the second individual personal characteristics, wherein the second personal characteristics value obtains from the second personal characteristics value of a plurality of multispectral images and corresponding collection is used second algorithm.

18. method according to claim 10, wherein, described algorithm is polynary algorithm.

19. according to the described method of one of claim 10 to 18, wherein, multispectral image is the spatial spectrum multispectral image.

20. a method of estimating individual personal characteristics, described method comprises:

Measure from individual collection of biological statistics; And

By application algorithm relation between biometric data measurement and personal characteristics value, determine the personal characteristics of individuality, described personal characteristics value obtains from the personal characteristics value of a plurality of biometric datas measurements and corresponding collection is used polynary algorithm.

21. method according to claim 20, wherein, one or more biometric data measurements comprise the measurement of skin line.

22. method according to claim 20, wherein, the measurement of collection of biological statistics comprises:

Utilize the skin part of rayed individuality;

Reception is from the light of skin part scattering; And

From the light that receives, obtain multispectral image.

23. according to claim 20 or 21 described methods, wherein, described biometric data measurement is that the spatial spectrum biometric data is measured.

24. method according to claim 22, wherein, described biometric data measurement is that the spatial spectrum biometric data is measured, and described multispectral image is the spatial spectrum multispectral image.

25. a sensing system comprises:

Illumination subsystems, being configured to provides light to the skin part of individuality;

Detection subsystem is configured to receive the light from the skin part scattering; And

Computing unit, join with detection subsystem, and have:

First instruction is used for obtaining multispectral image from receiving light; And

Second instruction is used for determining individual personal characteristics by use the algorithm relation between multispectral image and personal characteristics value that described personal characteristics value obtains from the personal characteristics value of a plurality of multispectral images and corresponding collection is used algorithm.

26. sensing system according to claim 25, wherein, illumination subsystems comprises:

Light source provides light in a plurality of discrete wave strong points; And

Illumination optics is used to make illumination to be mapped to skin part.

27. sensing system according to claim 26, wherein, described illumination subsystems also comprises scanning mechanism, and being used for according to the rules, pattern comes scan light.

28. sensing system according to claim 26, wherein, described light source comprises a plurality of quasi-monochromatic sources.

29. sensing system according to claim 25, wherein, described illumination subsystems comprises:

Wideband light source; And

Wave filter is provided for the light of filtering from the broadband light emitted.

30. sensing system according to claim 25, wherein, detection subsystem comprises:

Photodetector; And

Detection optics is used to make reception illumination to be mapped to photodetector.

31. sensing system according to claim 25, wherein:

Illumination subsystems comprises first polarizer, and the light in the face of providing is provided;

Detection subsystem comprises second polarizer, is provided for the light in the face of receiving; And

First and second polarizers are intersected with each other.

32. sensing system according to claim 25, wherein, computing unit also comprises the 3rd instruction, be used for by between the multispectral image and the second personal characteristics value, using second algorithm relation, determine the second individual personal characteristics, wherein the second personal characteristics value obtains from the second personal characteristics value of a plurality of multispectral images and corresponding collection is used second algorithm.

33. according to the described sensing system of one of claim 25 to 32, wherein, described multispectral image is the spatial spectrum multispectral image.

Images