Patent US 5,912,974

Patent №

US 5,912,974

Granted

Owner

Lab

AI components

1

vision

Assignment

None on record

Dataset

AIPD

2023_r1 edition

Application

08846818

Printed document contents are scanned and digitized, using a conventional scanner, and displayed page by page on a screen. Using a conventional editor and an input device, the scanned and digitized document contents are edited before being used to generate a digital signature. This allows reading errors, which could invalidate a subsequent verification process, to be corrected. Using the editor, and an input device such as a mouse, the signing authority identifies on the screen different segments of the document, each segment containing data of a single type and selects for each segment a set of rules, among a group proposed by the system, for authenticating it. Then, for each segment, an edited digital form of the data contents are derived using the method defined in the rules. A hash value of the rules used and the edited digital form of the segment contents is calculated using a public hashing algorithm. Then the apparatus generates a digital signature of the edited digitized segments contents using the secret key of the authenticator. Finally, an authentication code comprising the edited digital form of each segment and the digital signature is printed on the document. To verify the authenticity, the printed document is scanned and digitized again and the digital signature is checked using the associated public key. If the check fails, the verifier identifies which segment has been scanned differently, comparing it with the related edited digital form in the authentication code printed on the document to evaluate its validity.

VisionH04N 1/32133H04L 9/3247H04L 2209/60H04N 2201/3235H04N 2201/3236H04N 2201/3269H04N 2201/3271

AI classification

Vision0.99
AI hardware0.06
Evolutionary computation0.03
Knowledge representation0.03
Natural language0.00
Planning0.00
Machine learning0.00
Speech0.00
© 2026 NYSGPT2525 LLC