Dynamic Signature Verification (DSV)
DSV - Introduction
Most modern organizations today need a reliable means of personal identification. Unlike offline verification, dynamic verification of signatures uses the shape of an individual's signature and actually logs the micro timing of the hand movement ("choreography"). Consequently, it becomes more difficult to forge a dynamic signature.
The signature is recorded using a capture device, such as a digitizing tablet and a special pen. The capture device provides time-based parameters (coordinates, pressure, pen tilt, etc.) for each sampled point of the signature. This data is used for real time, automatic decisions, immediately after the person signs.
Figure 14 - On the left, a sample of a digitizer tablet, and on the right, a sample of signature matching
DSV - Problem Definition and Characteristics
The DSV problem is defined as follows: Given a large set of historical signatures (enrollment), of which N belong to a certain subject, and a tested signature, decide if the tested signature belongs to the specific subject (Accept) or it was forged (Reject).
False Acceptance (FA) and False Rejection (FR) should be minimized.
The DSV system developed by the Haifa team works in two modes:
- Enrollment of N (usually 6) signatures per subject into a database
- Verification of the tested signature against the enrolled bucket
Pros:
- The dynamic signature is a behavioral biometric measure.
- DSV is a socially and legally accepted procedure (in contrast to eye scans or fingerprints).
- Under normal conditions, people sign with relatively low variability.
Cons:
- Normal conditions (low variability) are often broken.
- People get nervous when asked to sign again.
- Signatures tend to change over the years.