Technical note: The Next Step - a semi-automatic coding and comparison system for forensic footwear impressions

Footwear impression evidence is a key tool in criminal investigations, connecting suspects to the crime scene; in addition, it may provide valuable forensic intelligence linking different crime scenes in the absence of a suspect. This paper presents the development of a computer database and semi-automatic system for shoeprint comparison. The database is based on structured manual coding of elements by CSI (Crime Scene Investigators). The computer algorithm then compares the pattern information of the query footwear impression to the entire database, returning a list of possible matches ranked by pattern similarity and crime scene geographic proximity. Initial results using a database of 284 footwear impressions from real crime scenes revealed that a "match" (i.e., the same shoe model) was found for 30% of the impressions; in most cases, the "match" ranked within the top five places of the "hit" list generated by the algorithm. Our results confirm that this semi-automatic footwear comparison system is simple, cost-effective and efficient, providing great potential for linking crime scenes.