A review of firearm toolmarks identification: Progress, challenges, and perspectives

Area of Science:

• Forensic Science
• Ballistics
• Imaging Technology

Background:

• Traditional firearm toolmark identification relied on 2D microscopy and subjective methods like Consecutive Matching Striae (CMS).
• This subjective approach faced judicial scrutiny due to a lack of statistical quantification and potential bias.

Purpose of the Study:

• To analyze the technological evolution of firearm toolmark identification from 2D to 3D methods.
• To highlight the shift towards probabilistic assessment frameworks for enhanced objectivity.

Main Methods:

• Utilized 3D microscopic imaging techniques (confocal microscopy, SEM, micro-CT).
• Employed deep learning and Congruent Matching Cells (CMC) for feature extraction.
• Applied Bayesian statistics to construct probabilistic models for likelihood ratio (LR) determination.

Main Results:

• 3D imaging and probabilistic methods provide quantifiable LR values, increasing objectivity.
• The shift enhances the credibility of firearm identification in legal proceedings.
• Current research focuses on automated systems and algorithm interpretability.

Conclusions:

• The evolution to 3D imaging and probabilistic analysis represents a significant advancement in firearm toolmark identification.
• Future research should address multimodal data fusion and the identification of non-standard firearms.
• The goal is to develop intelligent, standardized, and verifiable firearm identification systems.