Graphics Group @ ISU

Shoe analysis is frequently used as part of forensic analysis. However, dealing with time, the wear of the sole and its features can be difficult. Here we present a method for extracting the wear between two 3D scans of a single shoe after some amount of use, using an iterative alignment process, in order to discover matching methods after a large amount of wear. Initial processing transforms a Standard Triangle Language (stl) 3D object, into a triangle mesh before alignment. Alignment between the two meshes is accomplished using, principal component analysis to initially align the mesh objects onto the same plane, followed by an iterative closest point comparison. After alignment, surface differences are measured and compared, in order to define wear. This method has been used to successfully compare wear in a variety of brands including Nike, Adidas, and Sony for wear patterns from new up to 8 months old. The methods presented here allows for more accurate wear analysis which can help the forensic community match shoe print analysis, even after wear.