Overview
Shoe tracks are the evidences which are conserved the most by police forces on crime scenes. However it is very difficult to assign found shoe tracks automatically to reference tracks that are stored in a data base. We aim to build a new, smart, interactive system that helps to quickly assign unknown shoe tracks to reference tracks. This assignment process consists of two steps. In the first step, features from a shoe track indexing system are assigned to known reference tracks. In the second step, a user enters features into the system which are visible on the unknown track. The system is supporting the user by determining which features have the biggest expected information gain as well as returning the reference tracks ordered by their similarity to the stated search criteria.
