Spatial modeling of randomly acquired characteristics on outsoles with application to forensic shoeprint analysis

TL;DR

This paper models the distribution of randomly acquired characteristics on shoe soles as a point process, providing a statistical framework to improve forensic shoeprint analysis by identifying likely RAC locations.

Contribution

It introduces a novel point process model for RAC distribution on shoe soles, accounting for shoe variability and expert-defined regions, enhancing forensic comparison methods.

Findings

RACs tend to appear at specific locations related to foot morphology

The models identify high-probability RAC regions on shoe soles

Results can improve the evidential assessment in forensic investigations

Abstract

Footwear comparison is used to link between a suspect's shoe and a footprint found at a crime scene. Investigators compare the two items using randomly acquired characteristics (RACs), such as scratches or holes. However, to date, the distribution of RAC characteristics has not been investigated thoroughly, and the evidential value of RACs is yet to be explored. An important question concerns the distribution of the location of RACs on shoe soles, which can serve as a benchmark for comparison. The location of RACs is modeled here as a point process over the shoe sole and a data set of 386 independent shoes is used to estimate its rate function. The analysis is somewhat complicated as the shoes are differentiated by shape, level of wear and tear and contact surface. This paper presents methods that take into account these challenges, either by using natural cubic splines on high resolution data, or by using a piecewise-constant model on larger regions defined by experts' knowledge. It is shown that RACs are likely to appear at certain locations, corresponding to the foot's morphology. The results can guide investigators in determining the evidential value of footprint comparison.