초록 ▼

A dynamic signature/sign biometric verification system for detecting and preventing fraudulent transactions is described. The system comprises remote digital signature/sign input devices, a means to extract spatial and temporal features from the signature, a means to transmit the signature/sign feat

A dynamic signature/sign biometric verification system for detecting and preventing fraudulent transactions is described. The system comprises remote digital signature/sign input devices, a means to extract spatial and temporal features from the signature, a means to transmit the signature/sign features along with customer identifier information to a centralized signature/sign verification authority, a means for combining signature/sign feature verification with other forms of fraud detection technology, and a means for transmitting the results of a signature/sign verification back to the remote location where the signature/sign was captured. The system was primarily developed for use in payment card industries (e.g. credit cards, debit cards) but has applicability to other centralized signature/sign verification applications such as Automated Teller Machine authorizations and other identity theft detection and monitoring services.

대표청구항 ▼

1. A payment authorization system comprising: a) a first computer system and software comprising: i) an electronic payment processing system incorporating software; a method for capturing data from the purchaser's payment account; a method for capturing purchase transaction data; and a method of cap

1. A payment authorization system comprising: a) a first computer system and software comprising: i) an electronic payment processing system incorporating software; a method for capturing data from the purchaser's payment account; a method for capturing purchase transaction data; and a method of capturing an electronic copy of a purchaser's signature;ii) the capture of purchaser's signature data from an electronic signature entry system; the generation of signature feature set data from said purchaser's signature data;iii) the capture of said purchaser's payment account data, including said payment account number;iv) the capture of transaction data, including transaction amount and merchant identifier;v) the transmission of the said signature feature set data, said account data and said transaction data to said purchaser's payment account issuing bank as a purchase authorization request;vi) and the receipt from said purchaser's payment account issuing bank of the result of said purchase authorization request;b) a second computer system and software located at said purchaser's payment account issuing bank location that receives said purchase authorization request and i) automatically separates out said signature feature set data, said account data, and said transaction data into its constituent components;ii) automatically performs a signature match calculation by comparing said purchaser's signature feature set data with a previously stored signature template created from previous signature feature data sets from said purchaser; and generates one or more signature feature set data mismatch scores from said match calculation that is indicative of the likelihood that the signature is either authentic or fraudulent;iii) automatically analyzes said transaction data and said account data with pattern recognition software and/or fraud database analyses to identify high fraud risk transactions, and uses these analyses to generate one or more transaction fraud risk scores associated with the purchase request;iv) automatically screens account data to identify high credit risk transactions, including but not limited to verifying that said purchaser's payment account is in good standing and has sufficient purchasing capacity to approve said transaction amount;v) automatically approves said purchase authorization request responsive to: (1) said signature feature set data mismatch score being within a pre-determined acceptable range that would suggest the signature is authentic;(2) said transaction fraud risk score being within a pre-determined acceptable range that would suggest the transaction is not fraudulent; and(3) said account being in good standing and with sufficient purchasing capacity to support the purchase request;vi) either automatically or after human review, declines said purchase authorization request responsive to one or more of the following: (1) said signature feature set data mismatch score being outside of a pre-determined acceptable range that would suggest the signature may be a forgery;(2) said transaction fraud risk score being outside of a pre-determined acceptable range that would suggest the transaction may be fraudulent; or(3) said account not being in good standing;(4) said account having insufficient purchasing capacity to support the purchase request;vii) automatically transmits said purchase authorization request result to said first computer system. 2. The system of claim 1 where the electronic signature entry system is a Point-Of-Sale signature entry terminal. 3. The system of claim 1 where the electronic signature entry system originates as a signature area on a merchant's online store checkout screen and requires customer to sign said signature area using one of: a mouse, a stylus moving on mousepad, a finger moving on a mousepad, a finger moving on a touchscreen, a stylus moving on a touchscreen, and other computer pointing devices. 4. The system of claim 1 where the electronic signature entry system originates as a screen on a mobile device. 5. The system of claim 1 where the said transmissions are encrypted. 6. The system of claim 1 where the said signature feature set data has been developed from the purchaser's (X,Y,T) signature data. 7. The system of claim 1 where the said signature feature set data has been developed from the said purchaser's (X,Y,T) signature data after transforming said purchaser's signature to a consistent angle of inclination with respect to predetermined axes of an orthogonal coordinate positioning system, generating new data (x,y,t) from the said (X,Y,T) data. 8. The system of claim 7 where the said signature template contains mean estimates of the said biometric feature values calculated from different samples of the said signature feature set data. 9. The system of claim 8 where the said signature template contains estimates of deviations from the said biometric feature means calculated from different samples of said signature feature set data. 10. The system of claim 9 where the said match calculation measures the distance of each said biometric feature value from its said signature template mean value in relation to said signature biometric template estimate of deviation from said biometric feature mean estimate. 11. The system of claim 10 where the said biometric feature distances are modified with discriminating weights prior to being combined to produce at least one said signature feature set data mismatch score. 12. The system of claim 7 where the said signature feature set data includes functions of at least one of the following: a) V(x), where V(x) is the variance of the x-coordinate values of said transformed signature,b) V(y) where V(y) is the variance of the y-coordinate values of said transformed signature,c) C(x,y) where C(x,y) is the covariance of the (x,y) coordinate values of said transformed signature,d) Total signature time,e) Total in-contact signature time,f) Total out-of contact signature time,g) Positions of (x,y) turning points with respect to time,h) Positions of (x,y) turning points with respect to x-position,i) Positions of (x,y) turning points with respect to y-position,j) An estimate of total x-distance distance traveled,k) An estimate of total y-distance traveled,l) (x,y) positions of new points of stylus contact with respect to time,m) New out-of-contact stylus (x,y) positions with respect to x- position,n) (x,y) positions of new points of stylus contact with respect to x-positiono) New out-of-contact stylus (x,y) positions with respect to x-position,p) Forehand (x,y) distance, andq) Backhand (x,y) distances. 13. The system in claim 1 where the said capture of customer's signature represents sufficiently well-defined sample values, of sufficient quality to be represented by a recognizable printed signature constructed from them on a document containing data from the said account data and the said transaction data. 14. The system of claim 1 where the said signature biometric template is a dual purpose template and contains the said data representing the authentic signature of said purchaser as well as said signature feature set data. 15. The system in claim 1 where said signature feature set data is generated from said purchaser's signature, which is a sign, different from said authentic signature of said purchaser. 16. The system of claim 1 where said signature feature set data mismatch scores, said credit risk scores and said fraud risk scores are combined to arrive at said purchase authorization request result. 17. The system of claim 1 where a trusted third party performs the match calculation on said purchaser's signature feature set data using a previously stored signature template created from previous signature feature data sets from said purchaser. 18. A centralized signature verification authority system comprising: a) a computer system and software at a company or individual's location where a customer signs a document, comprising: i) a method to capture an electronic signature from a customer using an electronic signature entry system; the generation of a set of signature feature data from said signature;ii) the capture of a customer identification data set provided by a centralized signature verification authority;iii) the transmission of said signature feature data set and said customer identification data set to a centralized signature verification authority as a signature verification request; andiv) the receipt from said trusted centralized signature verification authority of the result of said centralized signature verification authority signature verification request;b) a second computer system and software located at said centralized signature verification authority's location, which is physically separated from the location where said customer signs said document, comprising: i) means for accepting said signature verification request;ii) means for matching the customer's identification data to an existing signature feature set template for that applicant stored in a database;iii) means for performing a match calculation from said signature feature set data received in the signature verification request with said existing signature feature set template for that customer;iv) means for determining whether said match calculation suggests that the customer's identity is authentic based on the signature feature set data comparison with said signature feature set template;v) means for combining the result of the signature feature set match with existing fraud screening methods in effect; andvi) means for providing the result of the said signature feature set data comparison to the company or individual requesting signature verification. 19. The system in claim 18 whereby the system is used as an identity theft prevention system that verifies the identities of new loan applicants prior to a bank approving a loan application. 20. The system in claim 18 whereby the system is used as an Automated Teller Machine (ATM) customer identity verification system.