A Multibiometric Approach in a Semi Automatic Dental Recognition Using DIFT Technique and Dental Shape Features

Teeth are one of the most important and popular biometric characteristic used in Forensic Dentistry. The development of automatic or semiautomatic, robust and precise forensic human identification systems are crucial for forensic, principally in mass disasters, like tsunamis and airplanes crashes. In this work we propose the use of the Differential Image Foresting Transform (DIFT) to extract the teeth and dental work contours from panoramic dental radiographs that are used as dental features. The shape descriptors based on the Shape Context method and Beam Angle Statistics (BAS) were implemented and evaluated for the teeth recognition. The Dental Code technique was used to evaluate all dental works, including the restorations. The experiments were carry out using a database of 1126 teeth images, obtained from 40 panoramic dental radiograph images from 20 individuals. The multibiometric approach improved the system performance generating, in the best case, an EER of 9%. KeywordsMultibiometrics; Forensic Dentistry; Dental Recognition; Image Forest Transform; Shape Context, Beam Angle Statistics, Edit Distance.