A database of handwriting samples for applications in forensic statistics

Area of Science:

• Forensic Science
• Biometrics
• Data Science

Background:

• Handwriting analysis is crucial in forensic investigations.
• Developing objective, statistical methods for handwriting analysis is an ongoing challenge.
• A need exists for large, well-documented datasets to train and validate such methods.

Purpose of the Study:

• To create a comprehensive dataset of adult handwriting samples.
• To facilitate the development of statistical approaches for handwriting analysis in forensic contexts.
• To provide a resource for researchers in forensic science and biometrics.

Main Methods:

• Collected handwriting samples from 90 adults across three sessions.
• Included survey data, demographic information, and session-specific details.
• Scanned writing samples and processed them into image files, ensuring data integrity through systematic management and an automated application.

Main Results:

• Generated a repository of 2430 handwriting sample images.
• Ensured data reliability via systematic document generation, QR code embedding, and automated data handling.
• Collected associated demographic and session-specific data for all participants.

Conclusions:

• The established dataset provides a valuable resource for advancing statistical methods in handwriting analysis.
• This work supports the objective evaluation of handwriting as forensic evidence.
• The data management strategies ensure the quality and usability of the forensic dataset.