Identity verification apparatus and method

Claim 3

The method of claim 1, also including the step of updating the reference vector if the handwriting sample being tested is accepted, by utilizing said accepted information vector to derive a new reference vector.

Claim 6

The method of claim 3, also including the step of updating the reference vector if verification of identity is indicated, by utilizing said accepted information vector to derive a new reference vector.

Claim 9

The method of claim 5 in which the step of deriving an information vector from the digital representation includes deriving a plurality of physical features in addition to the plurality of Haar coefficients.

Claim 10

The method of claim 6 in which said plurality of physical features includes a first group of features representative of the digital representation as a whole and a second group of features representative of each of a plurality of portions of said digital representation.

Claim 11

The method of claim 5 in which the comparing step includes determining whether the amount of deviation between the information vector for the writing of the person whose identity is being tested and the reference information vector exceeds a predetermined permissible amount.

Claim 12

The method of claim 5 in which the comparing step includes first determining the amount of deviation between said information vector and a reference vector, and also then comparing said amount of deviation with a predetermined deviation amount unique to said person.

Claim 15

The method of claim 10 in which the directly observable features include a first group of features representative of the compressed digital signal as a whole and a second group of features representative of each of a plurality of portions of said compressed digital signal.

Claim 16

The method of claim 10 in which the compressing step includes time-normalizing the signal to minimize the effect of those signal levels resulting from pen lift segments included in the act of providing the handwriting sample.

Claim 17

A machine-implemented method for deriving a standard for verifying the identity of a person from handwritten signatures, comprising the steps of A. providing a plurality of handwriting samples of the same informational content of a person for whom a reference information vector is desired; B. providing an electrical signal representative of each of said plurality of handwriting samples; C. deriving an individual signature vector from each one of said plurality of electrical signals; D. storing said signature vector; E. computing an average vector from said plurality of individual signature vectors; F. computing an individual error value for each signature vector based on a comparison of each signature vector with the average vector; G. computing an average error value using at least some of said individual error values; H. computing an acceptance error value based upon said average error value; I. comparing said individual error values with said acceptance error value; J. discarding any individual signature vector having an error value in excess of the acceptance error value; K. providing additional handwriting samples for deriving additional individual signature vectors to replace those discarded; L. computing individual error values for said additional individual signature vectors; M. comparing said additional individual error values with said acceptance error value and repeating steps A to J inclusive until a predetermined number of acceptable individual signature vectors have been obtained; N. computing a revised average signature vector from the acceptable individual signature vectors; O. computing revised individual error values for each acceptable individual signature vector based on a comparison of each revised individual signature vector with the revised average signature vector; P. computing a threshold value based upon the revised individual error values; Q. computing a normalizing vector assigning different weights to different aspects of the revised average information vector in accordance with the degrees of consistency of said different aspects in the various individual signature vectors; R. computing a reference vector based upon the revised average signature vector and the normalizing vector; and S. maintaining the threshold value and the reference vector to be utilized jointly as a standard for use in evaluating a vector derived from a handwritten signature of a person whose identity is to be verified.

Claim 18

A machine-implemented method for deriving a standard for verifying the identity of a person from handwritten signatures, comprising the steps of A. providing a plurality of handwriting samples of the same informational content of a person for whom a reference information vector is desired; B. providing an electrical signal representative of each of said plurality of handwriting samples; C. deriving an individual signature vector from each one of said plurality of electrical signals; D. storing said signature vector; E. computing an average vector from said plurality of individual signature vectors; F. computing an individual error value for each signature vector based on a comparison of each signature vector with the average vector; G. computing an average error value using at least some of said individual error values; H. computing an acceptance error value based upon said average error value; I. comparing said individual error values with said acceptance error value; J. discarding any individual signature vector having an error value in excess of the acceptance error value; K. providing additional handwriting samples for deriving additional individual signature vectors to replace those discarded; L. computing individual error values for said additional individual signature vectors; M. comparing said additional individual error values with said acceptance error value and repeating steps A to J inclusive until a predetermined number of acceptable individual signature vectors have been obtained; N. computing a revised average signature vector from the acceptable individual signature vectors; O. computing revised individual error values for each acceptable individual signature vector based on a comparison of each revised individual signature vector with the revised average signature vector; P. computing a threshold value based upon the revised individual error values; Q. computing a normalizing vector assigning different weights to different aspects of the revised average information vector in accordance with the degrees of consistency of said different aspects in the various individual signature vectors; R. computing a reference vector based upon the revised average signature vector and the normalizing vector; and S. maintaining the threshold value and the reference vector to be utilized jointly as a standard for use in evaluating a vector derived from a handwritten signature of a person whose identity is to be verified.

Claim 19

The method of claim 13 in which step C includes the steps of C1. digitizing said electrical signal; C2. compressing the digital signal; C3. time-normalizing the compressed digital signal; C4. extracting Haar coefficients and directly observable features from the compressed and time-normalized digital signal; and C5. deriving an individual signature vector from said Haar coefficients and said directly observable features.

Claim 20

A machine-implemented method for deriving a standard for verifying the identity of a person from samples of handwriting, comprrising the steps of A. providing a plurality of handwriting samples of the same informational content of a person for whom a reference information vector is desired; B. providing an electrical signal representative of each of said plurality of handwriting samples; C. deriving an individual information vector from each one of said plurality of electrical signals; D. computing an average information vector from said plurality of individual information vectors; E. computing individual error values and an average error value indicative of deviation of said individual information vectors from said avverage information vector; F. computing an acceptance error value based upon said average error value; G. discarding any individual information vector having an error value in excess of the acceptance error value; H. providing and testing additional individual information vectors until a predetermined number of acceptable individual information vectors have been obtained; I. computing a revised average information vector and revised individual error values from the acceptable individual information vectors; J. computing a threshold value based upon the revised individual error values; K. computing a normalizing vector which assigns different weights to different aspects of the revised average information vector; L. computing a reference vector based upon the revised average information vector and the normalizing vector; and M. providing a verification standard comprising the threshold value and the reference value to be used jointly in evaluating a vector derived from a handwriting sample of a person whose identity is to be verified.

Claim 21

A machine-implemented method for deriving a standard for verifying the identity of a person from samples of handwriting, comprrising the steps of A. providing a plurality of handwriting samples of the same informational content of a person for whom a reference information vector is desired; B. providing an electrical signal representative of each of said plurality of handwriting samples; C. deriving an individual information vector from each one of said plurality of electrical signals; D. computing an average information vector from said plurality of individual information vectors; E. computing individual error values and an average error value indicative of deviation of said individual information vectors from said avverage information vector; F. computing an acceptance error value based upon said average error value; G. discarding any individual information vector having an error value in excess of the acceptance error value; H. providing and testing additional individual information vectors until a predetermined number of acceptable individual information vectors have been obtained; I. computing a revised average information vector and revised individual error values from the acceptable individual information vectors; J. computing a threshold value based upon the revised individual error values; K. computing a normalizing vector which assigns different weights to different aspects of the revised average information vector; L. computing a reference vector based upon the revised average information vector and the normalizing vector; and M. providing a verification standard comprising the threshold value and the reference value to be used jointly in evaluating a vector derived from a handwriting sample of a person whose identity is to be verified.

Claim 22

The method of claim 15 in which step C includes the steps of C1. digitizing said analog signal; C2. compressing the digital signal; C3. extracting Haar coefficients and directly observable features from the compressed digital signal; and C4. deriving an individual information vector from said Haar coefficients and said directly observable features.

Claim 23

A machine-implemented method for deriving a standard for verifying the identity of a person from samples of handwriting, comprising the steps of A. providing a plurality of handwriting samples of the same information content of a person for whom a reference information vector is desired; B. providing an electrical signal representative of each of said plurality of handwriting samples; C. deriving an individual information vector from each one of said plurality of electrical signals; D. computing an average information vector from said plurality of individual information vectors; E. computing individual error values and an average error value indicative of deviation of said individual information vectors from said average information vector; F. computing an acceptance error value based upon said average error value; G. computing a threshold value based upon the individual error values; H. computing a normalizing vector which assigns different weights to different aspects of the average information vector; I. computing a reference vector based upon the average information vector and the normalizing vector; and J. providing a verification standard comprising the threshold value and the reference value to be used jointly in evaluating a vector derived from a handwriting sample of a person whose identity is to be verified.

Claim 24

A machine-implemented method for deriving a standard for verifying the identity of a person from samples of handwriting, comprising the steps of A. providing a plurality of handwriting samples of the same information content of a person for whom a reference information vector is desired; B. providing an electrical signal representative of each of said plurality of handwriting samples; C. deriving an individual information vector from each one of said plurality of electrical signals; D. computing an average information vector from said plurality of individual information vectors; E. computing individual error values and an average error value indicative of deviation of said individual information vectors from said average information vector; F. computing an acceptance error value based upon said average error value; G. computing a threshold value based upon the individual error values; H. computing a normalizing vector which assigns different weights to different aspects of the average information vector; I. computing a reference vector based upon the average information vector and the normalizing vector; and J. providing a verification standard comprising the threshold value and the reference value to be used jointly in evaluating a vector derived from a handwriting sample of a person whose identity is to be verified.

Claim 25

Apparatus for verifying the identity of a writer of handwritten information by comparing a signal generated from the writing of said information with a known reference, comprising manually operable means for inputting identification indicia to select a unique reference vector and a unique threshold which have been predetermined for the person whose identity is to be verified; pressure-responsive transducer means for converting forces generated from writing said information to an analog signal representative thereof; analog-to-digital converter means for generating a digital representation of said signal; means for deriving an information vector from said digital representation; means for determining the amount of deviation between said information vector and the reference vector for said person; comparing means for comparing the amount of deviation with the unique threshold for said person; and indicator means controlled by said comparing means for indicating verification of identity if the amount of deviation is less than the unique threshold and for indicating rejection of identity if the amount of deviation is greater than the unique threshold.

Claim 26

Apparatus for verifying the identity of a writer of handwritten information by comparing a signal generated from the writing of said information with a known reference, comprising manually operable means for inputting identification indicia to select a unique reference vector and a unique threshold which have been predetermined for the person whose identity is to be verified; pressure-responsive transducer means for converting forces generated from writing said information to an analog signal representative thereof; analog-to-digital converter means for generating a digital representation of said signal; means for deriving an information vector from said digital representation; means for determining the amount of deviation between said information vector and the reference vector for said person; comparing means for comparing the amount of deviation with the unique threshold for said person; and indicator means controlled by said comparing means for indicating verification of identity if the amount of deviation is less than the unique threshold and for indicating rejection of identity if the amount of deviation is greater than the unique threshold.

Claim 27

The apparatus of claim 18, also including means for updating the selected reference vector by recomputing said vector using the information vector derived from said handwritten information provided by the writer in the event that verification of identity is indicated.

Claim 28

Apparatus for verifying the identity of a writer of handwritten information by comparing a signal generated from the writing of said information with a known reference value, comprising pressure-responsive transducer means for converting forces generated from writing said information to an analog electrical signal representative thereof; analog-to-digital converter means for generating a digital representation of said signal; means for deriving an information vector from said digital representation including a plurality of Haar coefficients and a pluarlity of physical features; and means for comparing said information vector derived from said handwritten information with information derived from previous handwriting samples of the writer whose identity is to be verified to determine whether or not the identity of the writer is verified.

Claim 29

Apparatus for verifying the identity of a writer of handwritten information by comparing a signal generated from the writing of said information with a known reference value, comprising pressure-responsive transducer means for converting forces generated from writing said information to an analog electrical signal representative thereof; analog-to-digital converter means for generating a digital representation of said signal; means for deriving an information vector from said digital representation including a plurality of Haar coefficients and a pluarlity of physical features; and means for comparing said information vector derived from said handwritten information with information derived from previous handwriting samples of the writer whose identity is to be verified to determine whether or not the identity of the writer is verified.

Claim 30

The apparatus of claim 20 in which said plurality of physical features includes a first group of features representative of the digital representation as a whole and a second group of features representative of each of a plurality of portions of said digital representation.

Claim 31

The apparatus of claim 20 in which the comparing means includes error means for determining the permissible amount of deviation between the information vector for the writing of the person whose identity is being tested and the reference information vector.

Claim 32

The apparatus of claim 20 in which the comparing means includes first means for determining the amount of deviation between said information vector and a reference vector, and also includes second means for comparing said amount of deviation with a predetermined deviation amount unique to said person.

Claim 33

The apparatus of claim 20, also including means for updating said information derived from previous handwriting samples by recomputing said information using the information vector derived from said digital representation in the event that verification of identity is indicated.

Claim 34

Apparatus for deriving a standard for verifying the identity of a person from handwritten signatures, comprising means for deriving an individual signature vector from each one of a plurality of handwriting samples of the same informational content of a person for whom a reference information vector is desired; means for storing each said signature vector; means for computing an average vector from said plurality of individual signature vectors; means for computing an individual error value for each signature vector based on a comparison of each signature vector with the average vector; means for computing an average error value using at least some of said individual error values; means for computing an acceptance error value based upon said average error value; means for comparing said individual error values with said acceptance error value; means for discarding any individual signature vector having an error value in excess of the acceptance error value; means for requesting additional handwriting samples for deriving additional individual signature vectors to replace those discarded; means for computing individual error values for said additional individual signature vectors; means for comparing said additional individual error values with said acceptance error value; means for effecting the generation and testing of additional individual signature vectors until a predetermined number of acceptable individual signature vectors have been obtained; means for computing a revised average signature vector from the acceptable individual signature vectors; means for computing revised individual error values for each acceptable individual signature vector based on a comparison of each revised individual signature vector with the revised average signature vector; means for computing a threshold value based upon the revised individual error values; means for computing a normalizing vector assigning different weights to different aspects of the revised average signature vector in accordance with the degrees of consistency of said different aspects in the various individual signature vectors; means for computing a reference vector based upon the revised average signature vector and the normalizing vector; and means for maintaining the threshold value and the reference value to be utilized jointly as a standard for use in evaluating a vector derived from a handwritten signature of a person whose identity is to be verified.

Claim 35

Apparatus for deriving a standard for verifying the identity of a person from handwritten signatures, comprising means for deriving an individual signature vector from each one of a plurality of handwriting samples of the same informational content of a person for whom a reference information vector is desired; means for storing each said signature vector; means for computing an average vector from said plurality of individual signature vectors; means for computing an individual error value for each signature vector based on a comparison of each signature vector with the average vector; means for computing an average error value using at least some of said individual error values; means for computing an acceptance error value based upon said average error value; means for comparing said individual error values with said acceptance error value; means for discarding any individual signature vector having an error value in excess of the acceptance error value; means for requesting additional handwriting samples for deriving additional individual signature vectors to replace those discarded; means for computing individual error values for said additional individual signature vectors; means for comparing said additional individual error values with said acceptance error value; means for effecting the generation and testing of additional individual signature vectors until a predetermined number of acceptable individual signature vectors have been obtained; means for computing a revised average signature vector from the acceptable individual signature vectors; means for computing revised individual error values for each acceptable individual signature vector based on a comparison of each revised individual signature vector with the revised average signature vector; means for computing a threshold value based upon the revised individual error values; means for computing a normalizing vector assigning different weights to different aspects of the revised average signature vector in accordance with the degrees of consistency of said different aspects in the various individual signature vectors; means for computing a reference vector based upon the revised average signature vector and the normalizing vector; and means for maintaining the threshold value and the reference value to be utilized jointly as a standard for use in evaluating a vector derived from a handwritten signature of a person whose identity is to be verified.

Claim 36

The apparatus of claim 25 in which the means for deriving includes means for producing an analog electrical signal from forces measured during the writing of the sample; means for digitizing said analog signal; means for compressing the digital signal; means for extracting Haar coefficients and directly observable features from the compressed digital signal; and means for assembling an individual signature vector from said Haar coefficients and said directly observable features.

Claim 37

The apparatus of claim 26 in which the directly observable features include a first group of features representative of the compressed digital signal as a whole and a second group of features representative of each of a plurality of portions of said compressed digital signal.

Claim 38

Apparatus for deriving a standard for verifying the identity of a person from samples of handwriting, comprising means for deriving an individual information vector from each one of a plurality of handwriting samples of the same informational content of a person for whom a reference information vector is desired; means for computing an average information vector from said plurality of individual information vectors; means for computing individual error values and an average error value indicative of deviation of said individual information vectors from said average information vector; means for computing an acceptance error value based upon said average error value; means for computing a threshold value based upon the individual error values; means for computing a normalizing vector which assigns different weights to different aspects of the average information vector; means for computing a reference vector based upon the average information vector and the normalizing vector; and means for providing a verification standard comprising the threshold value and the reference value to be used jointly in evaluating a vector derived from a handwriting sample of a person whose identity is to be verified.

Claim 39

Apparatus for deriving a standard for verifying the identity of a person from samples of handwriting, comprising means for deriving an individual information vector from each one of a plurality of handwriting samples of the same informational content of a person for whom a reference information vector is desired; means for computing an average information vector from said plurality of individual information vectors; means for computing individual error values and an average error value indicative of deviation of said individual information vectors from said average information vector; means for computing an acceptance error value based upon said average error value; means for computing a threshold value based upon the individual error values; means for computing a normalizing vector which assigns different weights to different aspects of the average information vector; means for computing a reference vector based upon the average information vector and the normalizing vector; and means for providing a verification standard comprising the threshold value and the reference value to be used jointly in evaluating a vector derived from a handwriting sample of a person whose identity is to be verified.

Claim 40

Apparatus for deriving a standard for verifying the identity of a person from samples of handwriting, comprising means for deriving an individual information vector from each one of a plurality of handwriting samples of the same informational content of a person for whom a reference information vector is desired; means for computing an average information vector from said plurality of individual information vectors; means for computing individual error values and an average error value indicative of deviation of said individual information vectors from said average information vector; means for computing an acceptance error value based upon said average error value; means for discarding any individual information vector having an error value in excess of the acceptance error value; means for providing and testing additional individual information vectors until a predetermined number of acceptable individual information vectors have been obtained; means for computing a revised average information vector and revised individual error values from the acceptable individual information vectors; means for computing a threshold value based upon the revised individual error values; means for computing a normalizing vector which assigns different weights to different aspects of the revised average information vector; means for computing a reference vector based upon the revised average information vector and the normalizing vector; and means for providing a verification standard comprising the threshold value and the reference to be used jointly in evaluating a vector derived from a handwriting sample of a person whose identity is to be verified.

Claim 41

Apparatus for deriving a standard for verifying the identity of a person from samples of handwriting, comprising means for deriving an individual information vector from each one of a plurality of handwriting samples of the same informational content of a person for whom a reference information vector is desired; means for computing an average information vector from said plurality of individual information vectors; means for computing individual error values and an average error value indicative of deviation of said individual information vectors from said average information vector; means for computing an acceptance error value based upon said average error value; means for discarding any individual information vector having an error value in excess of the acceptance error value; means for providing and testing additional individual information vectors until a predetermined number of acceptable individual information vectors have been obtained; means for computing a revised average information vector and revised individual error values from the acceptable individual information vectors; means for computing a threshold value based upon the revised individual error values; means for computing a normalizing vector which assigns different weights to different aspects of the revised average information vector; means for computing a reference vector based upon the revised average information vector and the normalizing vector; and means for providing a verification standard comprising the threshold value and the reference to be used jointly in evaluating a vector derived from a handwriting sample of a person whose identity is to be verified.

Claim 42

The apparatus of claim 29 in which the means for deriving includes means for producing an analog electrical signal from forces measured during the writing of the sample; means for digitizing said analog signal; means for compressing the digital signal; means for extracting Haar coefficients and directly observable features from the compressed digital signal; and means for assembling an individual information vector from said Haar coefficients and said directly observable features.

Claim 43

The apparatus of claim 30 in which the directly observable features include a first group of features representative of the compressed digital signal as a whole and a second group of features representative of each of a plurality of portions of said compressed digital signal.

Claim 44

Apparatus for deriving an information vector for verifying the identity of a person from a handwriting sample, comprising transducer means for deriving an analog electrical signal from forces measured during the writing of the sample; means for digitizing said analog signal; means for compressing the digitized signal; means for extracting Haar coefficients and directly observable features from the compressed digital signal; and means for deriving an information vector from said Haar coefficients and said directly observable features.

Claim 45

Apparatus for deriving an information vector for verifying the identity of a person from a handwriting sample, comprising transducer means for deriving an analog electrical signal from forces measured during the writing of the sample; means for digitizing said analog signal; means for compressing the digitized signal; means for extracting Haar coefficients and directly observable features from the compressed digital signal; and means for deriving an information vector from said Haar coefficients and said directly observable features.

Claim 46

The apparatus of claim 32 in which the directly observable features include a first group of features representative of the compressed digital signal as a whole and a second group of features representative of each of a plurality of portions of said compressed digital signal.