Mitek gets US patent for fraudulent signature detection software

Source: Mitek Systems

Mitek Systems (BULLETIN BOARD: MITK), a leading provider of image analytics and pattern recognition software, has been awarded its fourth patent for its advancements in the accuracy, efficiency and security of signature validation and prevention of identity theft.

The software has been adopted for fraudulent signature detection on checks, voter ballots, account applications and other documents. Its unique signature encoding technology prevents signature theft by eliminating the need for an online library of authentic signatures.

How It Works

ImageNet Signatures encodes each target signature and compares it with encoded reference examples, rather than comparing actual signature images which may compromise privacy and security. Mitek's robust image analytics encode many different characteristics of each signature. The system accommodates natural variations in signatures by encoding multiple sample signatures.

The target signature can be captured from a digital pad, PDA, tablet PCs or a scanned paper document. In a banking environment, this flexibility enables signatures to be validated at any point of presentment.

"This fourth patent reaffirms Mitek's strong and growing intellectual property portfolio in the area of Image Analytics," says Mitek President and CEO Jim DeBello.

"Concerns about voter fraud and other forms of identity theft have made signature verification more important than ever before," says DeBello. "Mitek has incorporated this patented signature-encoding technology into its new ImageNet Signature developer's toolkit as well as its FraudProtect application, which has been designed to prevent signature images from being stolen by hackers as well as detect traditional signature forgery."

This same technology powers Mitek's SignProtect product, which is being used to validate physician signatures which are collected as proof that they received prescription drug samples provided by pharmaceutical firms.

Comments: (0)