System and method for check fraud detection using signature validation

Abstract

Systems and methods are provided for validating the authenticity of a signature on a document by providing a document from an account, the document including an actual signature and a machine-readable identifier, wherein the machine-readable identifier contains a string of data representing the integral characteristics of all valid account signatures and a person-specific confidence threshold. When the document is presented at a point of presentment, the document is scanned into a document-processing machine and the actual signature is compared against all valid account signatures.

Description

REFERENCE TO RELATED APPLICATIONS [0001] This application is a divisional application of U.S. patent application No. 11 / 009,251, filed on Dec. 10, 2004, the contents of which are incorporated herein by reference.FIELD OF THE INVENTION [0002] This invention relates to automated document processing, and more particularly to automatic processing of financial documents involving image-based signature verification. BACKGROUND OF THE INVENTION [0003] In general, financial institutions have automated most check processing systems by printing financial documents, such as account numbers and bank routing numbers onto the checks. Before a check amount is deducted from a payer's account, the amount, account number, and other important information must be extracted from the check. The highly automated form of extraction is done by a check processing control system that captures information from the Magnetic Ink Character Recognition (“MICR”) line. The MICR line consists of specially designed nu...

AI Technical Summary

Benefits of technology

[0019] A further aspect of the present invention involves encoding a document with account signature data for signature validation comprising compressing the account signature data into a fingerprint comprising machine-readable data corresponding to one or more account signatures such that the fingerprint contains only a small fraction of the account signature data, encoding the fingerprint into a barcode and printing the barcode on the document. The fingerprint represents integral characteristics of all valid account signatures and may identify a person-specific confidence threshold for each account user. According to some embodiments, the fingerprint comprises 100 bytes of data or less.

[0020] An additional aspect of the present invention involves encoding a document with account signature data for signature validation comprising compressing the account signature data to a fingerprint comprising machine-readable data corresponding to one or more account signatures, evaluating the complexity and topology of each account signature, encoding the fingerprint into a barcode and printing the barcode on the document. This aspect may further involve determining a person-specific confidence threshold for each account signature based upon the complexity and topology of each account signature.

[0021] Yet another aspect of the present invention involves locating one or more signatures on a check comprising receiving an original or preprocessed check bitmap, dividing the check bitmap into fixed-size tiles, sorting the tiles into predefined classes, identifying handwriting tiles, adjusting the signature positions using a connected components analysis and generating final signature locations. Sorting the tiles into predefined classes may include applying neural network classifiers to the tiles.

[0022] A further aspect of the present invention involves locating one or more signatures on a check, comprising receiving an original or preprocessed check bitmap having a predetermined signature default area, determining the location of one or more check layout elements, using the location of the one or more check layout elements to determine potential signature locations and employing a handwriting detection technique to detect the present or absence of handwriting in the potential signature locations.

Problems solved by technology

Check fraud is one of the largest challenges facing financial institutions today.

Although these check fraud security systems have been somewhat effective in deterring check fraud, they suffer from a multiplicity off drawbacks.

For example, these systems are generally very slow and prohibitively expensive.

One drawback of the above-identified signature verification systems is that they do not involve a compression of account signature data to a fingerprint containing only a small fraction of the account signature data.

These references also fail to disclose methods of determining person-specific confidence thresholds by evaluating the complexity and topology of account signatures.

A further drawback of the above-identified systems is that they assume a fixed-size signature representation.

Images