An effective numbering and classification system for dental panoramic radiographs

Abstract

In the current scenario, identifying a person in mass disasters is a challenging problem if adequate biometric information is not available. It can be handled by using dental radiographs as a suitable biometric means in such a circumstance. Forensic Odontology is a branch under biometrics which uses dental radiographs for human identification. In this paper an attempt is made to develop an algorithm for teeth numbering and classification, which is a major section in Automated Dental Identification system. Individual identification using dental panoramic images is an issue addressed in the literature. Hence, this work makes use of panoramic images for emerging this algorithm. The dental radiographs are initially preprocessed and each tooth is isolated for further processing. This image is subjected to Support Vector Machine classifier to segregate the tooth as Molar or Premolar. Then template matching algorithm followed by Universal numbering system is used to number the teeth. Experimental results show that this algorithm has numbering accuracy of 93.3% for Molar and 92% for premolar.