Abstract

Major disasters have highlightedthe significance of automated dental identification systems. For example in the 9/11 attacks, many victims were identifiable only from pieces of their jaw bones. This is the main motivation for building the Automated Dental Identification System (ADIS). The archiving and retrieving processes of dental records from large databases is a challenging task and has received inadequate attention intheliterature. This paper presents a new technique for retrieving dental records for ADIS. Our approach consists of two main stages. The first stage is the preprocessing stage of the dental records that includes segmentation and teeth labeling classificationin order to obtain reliable appearance-based, low computational-cost features. In the second stage we propose a technique based on using the Orthogonal Locality Preserving projection algorithm to produce a candidate list. The experiment results show that the proposed approach provides fast and effective results to search and retrieve a list of candidates from the database compared to other approaches suggested in the literature.

Key words: Dental Identification;Teeth Appearance Based;LaplacianteethSpaces, Candidate List; Orthogonal LocalityPreservingProjections (OLPP); Potentional Match Search Componet.