Surface granularity as a discriminating feature of illicit tablets

In this paper we propose an innovative methodology for automated profiling of illicit tablets by their surface granularity; a feature previously unexamined for this purpose. We make use of the tiny inconsistencies at the tablet surface, referred to as speckles, to generate a quantitative granularity profile of tablets. Euclidian distance is used as a measurement of (dis)similarity between granularity profiles. The frequency of observed distances is then modelled by kernel density estimation in order to generalize the observations and to calculate likelihood ratios (LRs). The resulting LRs are used to evaluate the potential of granularity profiles to differentiate between same-batch and different-batches tablets. Furthermore, we use the LRs as a similarity metric to refine database queries. We are able to derive reliable LRs within a scope that represent the true evidential value of the granularity feature. These metrics are used to refine candidate hit-lists form a database containing physical features of illicit tablets. We observe improved or identical ranking of candidate tablets in 87.5% of cases when granularity is considered.