In this paper, an approach for the identification of people based on digital orthopantomogram images is proposed and experimentally investigated. This approach is composed of four main stages. In the first stage, the image quality is enhanced using the Laplacian pyramid. In the second stage, the image is segmented into individual sub-images, each containing a single tooth. To do this, the line that separates the upper and lower jaw is obtained using integral projections, and then information about the intensity and location of particular types of tooth is applied. The extraction of the shapes of the teeth is the third stage. This stage also later involves each particular shape being represented using the Point Distance Histogram algorithm to obtain its description. Finally, the resultant descriptions are matched with the objects stored in a template base for a person and, using these, biometric identification is performed.
展开▼