Casualty Identification with Dental Radiographs and Photographs

Abstract

A biometric system is used to identify and authenticate a person based on the information, which are unique biological traits to work effectively. Physiological biometrics such as fingerprint, face, iris, etc., provide reliable information for human identification. However, most of these cannot be used in the case of identifying a person after a disaster. In this paper a dental biometric technique is presented based on dental radiographs and photographs. It is aimed to obtain casualty identification with photographs, in circumstances where radiographs are unavailable. Here, it is proposed to improve the shape extraction performance by using shape descriptors and skeleton. The obtained data from the test image are compared with the set of database images and the best match is obtained for authentication. An improved nodal-based matching is proposed with both dental radiographs and photographs to improve the matching rate. The final matching is carried out with both the shape information and nodal graphs obtained. The hit rate obtained is comparable with the existing algorithm with a rank −1 accuracy of 78%.