Victim identification with dental images using texture and morphological operations

Abstract

Victim identification in mass disasters is a pressing need at present. Since teeth are the most robust parts of the human body and also unique, they can be used as a key component if sufficient dental records are available. This research also aims at dental photographs if there is a nonavailability of radiographs in profusion. In order to reduce the search space, the teeth are initially classified as molar, premolar, and canine using a linear support vector machine and numbered using the universal numbering system. Since the postmortem records are detained for a long span after antemortem records, other issues like missing teeth are considered here. For individual identification, contour information has been proven as a useful measure so far. In addition to contour and skeleton, the texture information of an individual tooth is also incorporated in this paper. The prediction of contour and skeleton using angle and distance parameters is the novelty of this study. The cumulative matching characteristic curve and accuracy curves are plotted to validate the performance of this algorithm. The proposed method identifies the person accurately and signifies that it could give a substantial assistance to forensic odontologist during victim identification.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 10 citations exist upstream.