Automated comparison of cartridge cases using machine learning and statistical modelling

When a firearm is discharged, unique marks are imprinted on the cartridge case by various parts of the firearm (firing pin, breech face, extractor, and ejector). These marks can be used to identify the specific firearm that fired the cartridge. Forensic scientists use comparison microscopes to compare the markings on cartridge cases found at a crime scene with those test-fired from a suspect’s firearm. This can link a specific weapon to a particular crime. Cartridge cases found at different crime scenes can be compared to determine if the same firearm was used. This can help linking multiple crimes together and potentially identify serial offenders.

This project aims at developing machine learning algorithms, including deep learning, paired with statistical models to identify cartridge cases fired from the same gun to address the application above based on images. The machine learning method will be paired with statistical method to produce probabilistic classification methods that are amenable for forensic science.

Diversity statement

Our research community thrives on the diversity of students and staff which helps to make the University of Dundee a UK university of choice for postgraduate research.  We welcome applications from all talented individuals and are committed to widening access to those who have the ability and potential to benefit from higher education.