A Proposed Feature Extraction Technique for Dental X-Ray Images Based on Multiple Features

In recent years, the need for personal identification systems has increased in cases of severe accidents and calamities for criminal investigation. Biometric identification systems can be used for gaining access to systems and verification purpose. However, biometrics such as iris, fingerprint, etc. are vulnerable to early decay and decomposition. On the other hand, dental x-rays remain invariant over time and resist early decomposition, fire, etc. Hence they can be used for human identification purpose. In this paper, a novel approach to feature extraction of dental x-ray radiograph is proposed based on the shape and texture of extracted tooth from the radiograph and thereafter matching is done by finding mean square error between the query and database images.