Virtual and simulated striated toolmarks for forensic applications.

Area of Science:

• Forensic Science
• Materials Science
• Computational Modeling

Background:

• Forensic casework and research require numerous experimental toolmarks for analysis.
• Current methods are time-consuming, limiting the number of reproducible toolmarks.
• Accurate recovery of the angle of attack is crucial for toolmark examination.

Purpose of the Study:

• To develop a computational approach for predicting striated toolmarks from 3D screwdriver surface data.
• To validate virtual toolmarks against experimental marks and assess their similarity.
• To investigate toolmark similarity concerning the angle of attack and develop an automated method for its recovery.

Main Methods:

• Utilized 3D surface datasets of screwdrivers to generate virtual toolmarks.
• Compared virtual and experimental (wax) toolmarks qualitatively and quantitatively.
• Analyzed toolmark similarity at high angular resolution (1°) and employed wavelet analysis for structural study.

Main Results:

• Virtual toolmarks closely resemble real experimental marks.
• The developed framework successfully distinguishes between matching and non-matching toolmarks, even with angle of attack differences up to 40°.
• Automated angle of attack recovery achieved high accuracy and precision (0.618 ± 4.179°).

Conclusions:

• The computational approach significantly reduces the need for extensive experimental toolmark creation.
• The method provides a reliable framework for toolmark comparison and angle of attack determination in forensic science.
• Wavelet analysis offers insights into structural elements for realistic toolmark simulation.