Individualization of earprints

This article presents a study of the variation in the appearance of earprints from six pairs of monozygotic (MZ) twins, applying different methods with varying degrees of objectivity. By analyzing prints of MZ twins, we purposely select pairs of individuals for whom the degree of interindividual variation is low. We first provide an account of the corresponding characteristics and the differences that occur between prints from the individual members. We evaluate differences in both content (i.e., presence, shape, and intensity of imprinted features) and geometry (i.e., position of imprinted features). Differences are visualized by way of digital overlays. Differences in geometry are further evaluated by performing a hierarchical cluster analysis using variables derived from the position of seemingly corresponding features. As a next step, we explore a method to fully automatically analyze prints. In this approach, image regions are selected and matched automatically. A suggestion on how to apply this method to calculate the evidential value of an earprint is provided. To our knowledge, this is the first time that any method capable of fully automated earprint comparisons has been presented in the literature.