Enhancing Security on Touch-Screen Sensors with Augmented Handwritten Signatures

Area of Science:

• Biometrics and Human-Computer Interaction
• Cybersecurity and Personal Identity Verification
• Mobile Device Security

Background:

• Existing handwritten signature verification systems face vulnerabilities to attacks.
• Signature categorization based on data quality reveals varying security behaviors.
• Mobile touch-screen sensors require enhanced security measures for personal identity.

Purpose of the Study:

• To develop a novel strategy for reinforcing identity security on mobile touch-screen sensors.
• To enhance handwritten signature resistance to attacks by augmenting information content.
• To assess the impact of augmented signatures on data quality and verification performance across different signature categories.

Main Methods:

• Augmenting handwritten signatures with additional personal handwritten inputs (alphanumeric, handwriting) at the enrollment phase.
• Utilizing clustering techniques with an entropy-based data quality measure to categorize signatures.
• Evaluating the performance of augmented signatures on different user categories acquired in uncontrolled mobile conditions.
• Assessing security enhancements per signature category, focusing on data quality and verification performance.

Main Results:

• The study analyzed signature types from 74 users on a widely deployed mobile touch-screen sensor.
• Adding alphanumeric (date) and handwriting (place) information proved to be the most effective augmentation strategy.
• This augmented signature type demonstrated significant improvements in verification performance across all user categories.
• A relative improvement of at least 93% in verification performance was observed compared to standard signatures.

Conclusions:

• Augmenting handwritten signatures with specific personal details is a powerful strategy to enhance mobile identity security.
• The proposed sensor-level enrichment significantly improves signature resistance to attacks and verification accuracy.
• The combination of alphanumeric and handwriting data offers superior security compared to traditional signatures, especially in uncontrolled mobile environments.