WO1998024051A1 - Procede de verification d'une signature echantillon avec une signature de reference choisie - Google Patents

Procede de verification d'une signature echantillon avec une signature de reference choisie Download PDF

Info

Publication number
WO1998024051A1
WO1998024051A1 PCT/DE1997/002767 DE9702767W WO9824051A1 WO 1998024051 A1 WO1998024051 A1 WO 1998024051A1 DE 9702767 W DE9702767 W DE 9702767W WO 9824051 A1 WO9824051 A1 WO 9824051A1
Authority
WO
WIPO (PCT)
Prior art keywords
stroke
sample
assignment
strokes
lettering
Prior art date
Application number
PCT/DE1997/002767
Other languages
German (de)
English (en)
Inventor
Brigitte Wirtz
Original Assignee
Siemens Aktiengesellschaft
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
Application filed by Siemens Aktiengesellschaft filed Critical Siemens Aktiengesellschaft
Publication of WO1998024051A1 publication Critical patent/WO1998024051A1/fr

Links

Classifications

    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
    • G07C9/00—Individual registration on entry or exit
    • G07C9/30—Individual registration on entry or exit not involving the use of a pass
    • G07C9/32—Individual registration on entry or exit not involving the use of a pass in combination with an identity check
    • G07C9/35—Individual registration on entry or exit not involving the use of a pass in combination with an identity check by means of a handwritten signature
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/30—Writer recognition; Reading and verifying signatures

Definitions

  • the invention relates to a method for verifying a sample lettering with any reference letter.
  • Physiological methods use human characteristics that normally do not change.
  • biometric features are that, in principle, they cannot be stolen and can only be copied with difficulty.
  • password or chip card systems it can be checked whether the card or the key is valid, but it is not checked whether the current user is also the legal owner of the respective means of identification.
  • Signature verification is one such biometric process.
  • a signature verification system processes the signature - regardless of the text content - with the aim of either confirming or rejecting the identity of the writer.
  • [1] describes a method for dynamic verification of a lettering using a reference lettering.
  • a commercially available electromagnetic tablet is used to record the signature, which records as data whether the stroke is carried out on (tablet stroke) or over (close range stroke) the tablet surface, the coordinates (along the x or y axis), the contact pressure and a discrete sequence of sampling times.
  • the sequence of tablet strokes that is to say pen movements carried out on the tablet surface, and close-range strokes, that is to say pen movements carried out in the vicinity of the tablet surface, are referred to in a signature signal by stroke structure. From the procedure known under [1], the signer does not take into account changes in the signature, that is, changes in the stroke structure.
  • [3] discloses a method for determining a reference lettering on the basis of a number of model-identical lettering types.
  • [4] specifies a method and an arrangement for improving a signature verification in order to reduce the number of incorrectly rejected signatures.
  • [5] deals with a procedure for real-time signature verification.
  • signals that describe the signature electronically are summarized and evaluated in order to ensure stable verification and consequently a low rate of incorrect rejections.
  • the object of the method according to the invention is to suitably evaluate pen withdrawals from and deletions of the pen on the surface which the signer has unconsciously carried out.
  • the method according to the invention verifies one
  • Sample lettering based on any reference lettering both the sample lettering and the Reference lettering has a stroke structure each consisting of at least one stroke.
  • the entire sample lettering as well as reference lettering consists of a series of sample strokes or reference strokes.
  • a distance measure can be determined, for example, by means of dynamic programming [2] between the sample stroke and the reference stroke. If the respective sample stroke considered is a short-range stroke, a 1: 1 assignment can be concluded and the next sample stroke can be continued directly. It is also possible to proceed directly to the next sample stroke if the calculated distance measure is within an allowable range and the respective sample stroke is therefore considered to be suitable for the respective reference stroke. In this case, too, one speaks of a 1: 1 assignment, because exactly one sample stroke is assigned to a reference stroke.
  • an assignment measure can be defined that combines the accumulation of all distance measures.
  • Each decision for an assignment can be associated with a differently determined distance measure, which is added to the assignment measure after each assignment. If the entire sample lettering with all of its sample strokes has been processed, the allocation dimension contains, accumulated, all the distance dimensions that result from the allocation made. It can therefore be decided on the basis of the value of the assignment measure whether the sample lettering is authenticated on the basis of the reference lettering.
  • the distance measure exceeds a predeterminable
  • Threshold different distance dimensions are calculated according to a connection rule.
  • the connection rule is used to successively combine both sample strokes and reference strokes according to an S [NS] * rule, which states that by successively adding ⁇ short-range stroke, tray stroke ⁇ pairs to either of the shorter ones Pattern stroke or reference stroke is continued until the lengths of the strings strung together are sufficiently similar, and alternative assignments of at least one sample stroke to at least one reference stroke are sought.
  • the method according to the invention takes into account different distance dimensions in the event of a stroke being omitted, for the case of an additional stroke, for the case of varying stroke structures and for the case of a 1: 1 assignment. To prevent the connection rule due to bad but necessary assignments
  • Sample strokes and / or reference strokes can be summarized as far as desired, although the 1: 1 assignment would have been correct, shortening rules can also be introduced.
  • the shortening rules are intended to prevent, for example, counterfeiters with usually poorer assignment values for the individual sample strokes from being combined as much as desired - in extreme cases the entire signature into one unit - due to the higher distance dimensions for 1: 1 assignments, and thus the assignment measure by stringing sample strokes together " is artificially improved.
  • Shortening rule can thus prevent a combination of sample strokes and / or reference strokes from being permitted only if this is "significantly" better than the 1: 1 assignment. Furthermore, a series of a few sample strokes and / or
  • the appropriate distance measure can be determined from the different distance measures using a minimum rule by selecting the most likely distance measure.
  • the distance measure determined can be added to the assignment measure, and according to the assignment made, the next sample stroke is determined for a further iteration of the method according to the invention, as long as both the sample lettering and the reference lettering have not yet been processed. It should be noted that if the minimum rule is on Omission of a stroke or an additional stroke decides, as the next sample stroke / reference stroke, the following tray stroke is examined.
  • connection rule calculates a distance measure for a varying stroke structure with a suitable aspect ratio between sample stroke and reference stroke.
  • the shortening rule is applied in such a way that a measure of movement, that is to say a measure of the dynamics during the signing, is decisive for the assignment, regardless of the length of the series of sample strokes.
  • the 1: 1 assignment can be made for the shortening rule if it has a similar status measure to the other agreements explained in the shortening rule.
  • the minimum rule selects the appropriate one from the alternative assignments determined by the connection rule.
  • the distance measure for the omission of a stroke or the distance measure for an additional stroke is only insignificantly smaller than the distance measure for the 1: 1 assignment or the distance measure for varying stroke structures, the more likely variant is selected.
  • the 1: 1 assignment is selected anyway if the distance measure of the subsequent assignments of sample strokes is only sufficiently similar to the minimum distance measure.
  • the 1: 1 assignment of an S [NS] * assignment with a small distance dimension is preferred if shape and movement are sufficient as criteria in the comparison of sample stroke with reference stroke.
  • An advantageous form of the method according to the invention can be the classification of individual points in the signature by means of handwriting recognition methods. This way, points can be explicitly recognized and their occurrence either generally or for certain writers at any point in the signature.
  • FIG. 1 shows a sketch with different signatures, all of which are authentic and yet vary in their stroke structure
  • FIG. 2 shows a sketch in which alternating stroke structures of an authentic signature are shown
  • FIG. 3 shows a block diagram that represents the method according to the invention.
  • Fig.l shows real signature samples and shows how the stroke structure of original signatures can vary while maintaining the overall impression of the signature.
  • Varying stroke structures can be caused, for example, by unintentional lifting of the pin
  • the solid lines of tray strokes indicate the dotted lines of near-range strokes.
  • the stroke structure is ⁇ Ri, i-Punkt, e, g, el, T, h, om, a, s ⁇ , in (b) ⁇ Ri, i-Punkt, eg, el, T, h, oma, s ⁇ , in (c) ⁇ Ri, i-Punkt, e, g, el, T, h, om, a, s ⁇ and in (d) ⁇ Ri, i-Punkt, eg, el, T, h, additional "a", om, a, s ⁇ .
  • Fig.l the stroke structure is ⁇ Ri, i-Punkt, e, g, el, T, h, om, a, s ⁇ .
  • Example signatures differ mainly in the stroke structure, but not in their overall appearance, illustrated here by the position data. Further data dependent on the dynamics, such as the contact pressure during the discrete sampling times, are not taken into account in FIG. Numerous test runs have shown that the signer alternately carries certain parts of the signature both on (tablet strokes) and above the tablet surface (close-range strokes) without being aware of it. If, for example, the signature were (a) the reference, (b) the sample signature, "e” or “g” would be mapped from the reference to “eg” in the sample, depending on the better distance dimension, and thus "e” or “g” as missing in the sample.
  • the basic sequence of the method according to the invention can be represented as follows: It is assumed that the 1: 1 assignment (a sample stroke is mapped to a reference stroke or vice versa) is normally the correct one, the corresponding strokes are not compared linearly and the distance measures obtained are accumulated into an assignment measure. Alternative assignments are only possible if the distance between and the length ratios of the reference stroke and sample stroke differ too much calculated. In general, the most likely alternative is the existence of a 1: 1 assignment, the second most likely alternative is the varying stroke structure and finally, as the third most likely alternative, it is either the omission of a stroke or an additional stroke in the
  • connection rule searches for the best distance dimension between at least one sample stroke and at least one reference stroke by successively combining strokes according to the S [NS] * rule.
  • shortening rules ensure that, because of bad but necessary assignments, the grouping is not extended to wherever the 1: 1 assignment would have been the correct one. For example, in the case of counterfeiters with usually poorer spacing dimensions for the individual strokes, the poor spacing dimensions for individual stroke assignments can be combined as much as desired - in extreme cases the entire signature into one unit - and thus the forged signature is verified.
  • the shortening rules are intended to prevent artificially improved distance dimensions from being obtained by combinations of strokes. A summary is only permitted if this
  • Fig. 2 illustrates the situation of changing stroke structures.
  • a reference lettering REF given, above and below each a sample lettering MU1 and MU2 with stroke structures that differ from the reference lettering REF.
  • the three black dots in the respective lettering indicate the short-range stroke.
  • the vertical dashed lines show the assignments of the respective sample strokes to the reference strokes based on the sum of the stroke lengths.
  • Tray stroke "T” a 1: 1 assignment of sample stroke to reference stroke instead.
  • the subsequent short-range stroke is also assigned to the associated reference short-range stroke.
  • the connection rule is applied in such a way that three sample strokes are assigned to a reference stroke. A comparison of the remaining sample strokes results in a 1: 1 assignment to the respective reference strokes.
  • the sample lettering MU2 can again be displayed 1: 1 on the reference lettering REF for the first two sample strokes.
  • the 1: 1 assignment fails for the remaining three strokes in the MU2 sample lettering.
  • the connection rule is applied gradually. Finally, it follows that only the sequence of the three remaining sample strokes can be compared with the sequence of the three reference strokes.
  • a sample lettering is to be verified using any reference lettering. Both the sample lettering and the reference lettering are each based on at least one stroke.
  • the method begins by comparing the first sample stroke with the first reference stroke. A distance measure is used for this D determined for comparison of the sample stroke with the reference stroke (step 3a). The distance dimension D is calculated using dynamic programming [2] as an example. If the current sample stroke is a short-range stroke (decision in step 3b), the calculated distance measure is added to an assignment measure that was set to 0 at the start of the verification (step 3c). The next sample stroke and the next reference stroke are considered (step 3d) and, after checking whether all strokes have been processed (step 3e), jumped to step 3a.
  • step 3b it is examined whether the distance between the sample stroke and the reference stroke is in a predefinable range and the difference in the stroke lengths of the sample stroke and the reference stroke is sufficiently similar (see step 3b) . If this is the case, the 1: 1 assignment is made. In the 1: 1 assignment, the distance dimension D is added to the assignment dimension ZM (step 3c) and jumped to the next stroke both in the sample lettering and in the reference lettering (step 3d). If all strokes have been processed both in the sample lettering and in the reference lettering (step 3e), the method terminates (step 3f). If there are still unprocessed strokes both in the pattern lettering and in the reference lettering, the process jumps again to step 3a, where the distance dimension D is calculated between the newly selected strokes.
  • connection rule is applied.
  • the connection rule searches for successive combinations of strokes according to the S [NS] * rule for alternative assignments of at least one reference stroke to at least one sample stroke. It is determined whether the length ratio between the sample stroke and the reference stroke is within a predefinable limit (see step 3g). If this is guaranteed, the connection rule is applied in such a way that a distance measure is calculated in accordance with the alternatives for omitting a stroke in the sample lettering, for an additional stroke in the sample lettering and for a 1: 1 assignment (step 3i). If the length ratio mentioned is outside the predefinable limit, in addition to the alternatives just mentioned, the alternative for varying stroke structures is also taken into account and a corresponding distance measure is calculated (step 3h).
  • connection rule shortening rules ensure that rows of strokes of any length are not combined due to poor but necessary assignments, although the 1: 1 assignment would have been the correct one. If the spacing dimensions obtained are bad in an assignment, but the ratio of the stroke lengths from sample stroke to reference stroke is appropriate, it must be checked whether it is a correct (but bad) or a missing (omission) or additional stroke. The alternative assignments for omission or addition must be determined using the connection rule. If the length ratios of reference stroke to sample stroke do not match, it can be a "bad correct" one.
  • the distance measure is above all a different stroke structure. So three alternatives must be checked using the connection rule.
  • step 3j The appropriate alternative assignment is selected using the minimum rule (step 3j).
  • step 3k a corresponding assignment is made in step 3k, the distance measure is added to the assignment measure (step 31) and the next relevant stroke is jumped to both in the reference lettering and in the sample lettering taking into account the assignment (step 3m ).
  • the minimum rule decides on an additional stroke, the distance dimension D- ⁇ is added to the allocation dimension ZM and in the sample lettering as well as in the reference lettering jump to the next relevant stroke taking into account an additional stroke (additional tray stroke means additional pair of near-range stroke and Tray stroke). In the latter case, the minimum rule decides to omit a stroke, the distance dimension D 2 is added to the allocation dimension ZM and it is both in
  • step 3e a comparison is made as to whether all strokes have been processed both in the sample lettering and in the reference lettering. If this is the case, the procedure terminates (step 3f). If all strokes have not yet been processed, the process branches back to the point at which the distance dimension D is calculated (step 3a).
  • connection rule occurring in the method according to the invention is defined as follows: (VI) New test combinations for the application of the connection rule are created by appending a ⁇ close-range stroke, tray stroke ⁇ pair to the previously shorter stroke series. (V2) Only strokes are considered local
  • connection rule The following shortening rules are used in the connection rule:
  • Stroke combination i.e. omission or additional strokes
  • M2 the distance dimensions of the 1: 1 assignment are significantly higher than the minimum
  • the 1: 1 assignment is preferred if the distance dimensions of the successor assignment, i.e. the tray stroke assignment following the 1: 1 assignment, are sufficiently similar to those of the variant with minimal distance dimensions. If the following assignments fit again, it can be assumed that the 1: 1 assignment was bad but correct.
  • the method according to the invention is based on the comparison of strokes as the underlying writing units.
  • the advantage of using strokes is that they represent a robust pre-segmentation of the signature and the font segments, as they are obtained from a digitizing tablet, can be used directly without the need for further data pretreatment.
  • the procedure is without limitation to any other
  • An advantageous development of the method according to the invention consists in explicitly recognizing points such as i-points with the help of handwriting recognition methods and their occurrence either generally or for certain writers at any place on the signature to allow. This "i-point rule" is then also an alternative assignment to be tested.
  • pen-dependent signature characteristics as personal rules for the respective user in the method according to the invention.
  • An example is the recognition of the fact that an i-point can be set at different points in the stroke order.
  • Comparative dimensions compared to the length ratios weighted sp_test beta * sp the test standard deviation for stroke lengths dp the function that compares and assigns signal elements, here signature parts, using dynamic programming, and oBdA at least one dimension m_shape for shape, one dimension m_motion for movement and that Assignment function warp delivers.
  • sim_shape 0.6 measure for sufficient similarity in
  • sim_motion 0.7 measure for sufficient similarity in
  • connection rule as part of the method according to the invention searches for an assignment from a reference stroke i to a sample stroke j from a touchdown point for alternative assignments of at least one reference stroke to at least one sample stroke by successively combining strokes according to the S [NS] * rule.
  • connection rule is called connect_strokes, i and j name the starting points for checking alternative assignments in the reference lettering and in the sample lettering. There are different calls to the connection rule: connect_strokes (i, j) test for changed stroke structure in the current assignment point, connect_strokes (i, j + 2) test for additional stroke on the
  • delta 0.46 threshold for length differences between reference and pattern, above which no alternative assignments are tested.
  • connection rule connect_strokes in pseudo code notation is agreed as follows:
  • MSort MSort (ISort),

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Character Discrimination (AREA)
  • Collating Specific Patterns (AREA)

Abstract

Le procédé permet de vérifier une signature échantillon au moyen d'une signature de référence, les deux signatures présentant chacune une structure de trait comprenant au moins un trait. Grâce au procédé selon l'invention, des soulèvements ou des abaissements, même involontaires, de la pointe d'écriture, comparativement à la signature de référence, peuvent être interprétés correctement pendant la signature, en tenant compte aussi bien des traits sur la surface d'une tablette électronique, que des traits au-dessus de cette surface. Des rapports de longueur entre des traits de l'échantillon et des traits de référence, ainsi que des espacements différents, sont calculés, au moyen d'une règle de liaison, dans le cas de l'omission d'un trait, d'un trait supplémentaire, d'une structure de trait variable ou encore, d'une correspondance 1:1. Des règles de contraction dans la règle de liaison permettent d'examiner s'il n'est pas choisi une trop longue série de traits les uns à la suite des autres. La décision pour une variante de correspondance est fournie par une règle du minimum permettant d'extraire la correspondance la plus probable.
PCT/DE1997/002767 1996-11-29 1997-11-26 Procede de verification d'une signature echantillon avec une signature de reference choisie WO1998024051A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE19649692.6 1996-11-29
DE19649692A DE19649692C1 (de) 1996-11-29 1996-11-29 Verfahren zur Verifikation eines Musterschriftzuges mit Hilfe eines Referenzschriftzugs

Publications (1)

Publication Number Publication Date
WO1998024051A1 true WO1998024051A1 (fr) 1998-06-04

Family

ID=7813238

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/DE1997/002767 WO1998024051A1 (fr) 1996-11-29 1997-11-26 Procede de verification d'une signature echantillon avec une signature de reference choisie

Country Status (2)

Country Link
DE (1) DE19649692C1 (fr)
WO (1) WO1998024051A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE10132012B4 (de) * 2001-07-03 2004-02-12 Siemens Ag Skalierung von Ähnlichkeitsmaßen für ein biometrisches Verfahren
DE10132013B4 (de) * 2001-07-03 2004-04-08 Siemens Ag Multimodale Biometrie
CN104077268A (zh) * 2013-03-25 2014-10-01 株式会社东芝 整形装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4561105A (en) * 1983-01-19 1985-12-24 Communication Intelligence Corporation Complex pattern recognition method and system
WO1987006744A1 (fr) * 1986-04-28 1987-11-05 Eric Rothfjell Procede de verification d'une signature et dispositif pour sa mise en oeuvre
EP0391044A2 (fr) * 1989-04-03 1990-10-10 Cadix Inc. Méthode de génération de modèles pour enregistrer une écriture manuscrite
WO1995016974A1 (fr) * 1993-12-17 1995-06-22 Quintet, Incorporated Procede de verification automatisee de signature

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5109426A (en) * 1989-11-10 1992-04-28 National Research Development Corporation Methods and apparatus for signature verification
DE19511472C1 (de) * 1995-03-29 1996-10-17 Siemens Ag Verfahren zur dynamischen Verifikation eines Schriftzuges anhand eines Referenzschriftzuges
DE19511470C1 (de) * 1995-03-29 1996-10-24 Siemens Ag Verfahren zur Ermittlung eines Referenzschriftzuges anhand einer Menge von schreiberidentischen Musterschriftzügen

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4561105A (en) * 1983-01-19 1985-12-24 Communication Intelligence Corporation Complex pattern recognition method and system
WO1987006744A1 (fr) * 1986-04-28 1987-11-05 Eric Rothfjell Procede de verification d'une signature et dispositif pour sa mise en oeuvre
EP0391044A2 (fr) * 1989-04-03 1990-10-10 Cadix Inc. Méthode de génération de modèles pour enregistrer une écriture manuscrite
WO1995016974A1 (fr) * 1993-12-17 1995-06-22 Quintet, Incorporated Procede de verification automatisee de signature

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE10132012B4 (de) * 2001-07-03 2004-02-12 Siemens Ag Skalierung von Ähnlichkeitsmaßen für ein biometrisches Verfahren
DE10132013B4 (de) * 2001-07-03 2004-04-08 Siemens Ag Multimodale Biometrie
US7092553B2 (en) 2001-07-03 2006-08-15 Siemens Aktiengesellschaft Multimodal biometry
CN104077268A (zh) * 2013-03-25 2014-10-01 株式会社东芝 整形装置

Also Published As

Publication number Publication date
DE19649692C1 (de) 1998-06-10

Similar Documents

Publication Publication Date Title
DE19547812C2 (de) Lesegerät für Schriftzeichenketten
DE69827177T2 (de) System zum Vergleich von Streifenmuster
EP2102794B1 (fr) Support de données portable pour la détection biométrique d'utilisateurs
DE102006057552A1 (de) System und Verfahren zur Messung des Abstands eines vorausfahrenden Fahrzeugs
DE2740394A1 (de) Automatisches vergleichssystem fuer merkmale eines musters
DE19511470C1 (de) Verfahren zur Ermittlung eines Referenzschriftzuges anhand einer Menge von schreiberidentischen Musterschriftzügen
EP0980565B1 (fr) Procede d'adaptation par ordinateur d'un jeu de donnees de reference a l'aide d'un jeu de donnees d'entree
DE60200829T2 (de) Vorrichtung und Verfahren zur biometrischen Überprüfung und Erfassung der Identität einer Person mittels Fingerabdruckinformation
EP0964390A2 (fr) Dispositif pour la vérification de signaux
DE19511472C1 (de) Verfahren zur dynamischen Verifikation eines Schriftzuges anhand eines Referenzschriftzuges
EP1897067A1 (fr) Procede et dispositif de detection d'une monnaie par utilisation de son image estampee
DE10260642B4 (de) Verfahren zur Erstellung einer Referenz für Fingerprints und zum Vergleich von Fingerprints
DE19649692C1 (de) Verfahren zur Verifikation eines Musterschriftzuges mit Hilfe eines Referenzschriftzugs
CH717006A2 (de) Verfahren zur Benutzeridentifikation.
DE60100757T2 (de) Verfahren und einrichtung zur verarbeitung von bildern
DE60214033T2 (de) Verfahren und einrichtung zum vergleichen von fingerabdrücken
EP2077658A1 (fr) Procédé de mise à disposition d'un service pour un utilisateur
DE60019136T2 (de) Verfahren und vorrichtung zur aufnahme und verifikation von fingerabdruck-informationen
EP1402460B1 (fr) Identification et/ou verifications automatiques de textures a lignes telles que des empreintes digitales
DE69734507T2 (de) Verfahren zur optimierung eines erkennungswörterbuchs, so das ähnliche muster besser unterschieden werden können
DE102005015180A1 (de) Verifizieren von Fingerabdrücken
EP0843864B1 (fr) Procede de classification et d'identification de modeles selon lequel une signature est produite par lissage d'un contour polygonal
DE69725252T2 (de) Verfahren und Vorrichtung zur Prüfung von Sprache
DE10206658A1 (de) Verfahren zum Überprüfen einer integrierten elektrischen Schaltung
DE10132012B4 (de) Skalierung von Ähnlichkeitsmaßen für ein biometrisches Verfahren

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): US

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
WWE Wipo information: entry into national phase

Ref document number: 1997951099

Country of ref document: EP

WWW Wipo information: withdrawn in national office

Ref document number: 1997951099

Country of ref document: EP

122 Ep: pct application non-entry in european phase