Person identification using SURF features of dental radiograph

This paper presents a model that is designed to identify people based on shapes and appearances of their teeth. In the proposed model, wiener filtering is used to reduce noise and provide smooth image. This filtered image is used in canny edge detection algorithm where image's edge is defined through Gaussian Filtering and edge thresholding which determines the region of the shape that represents binary pixel value. This pixel value can be used in image identification. Furthermore, SURF algorithm is used to define interest points. Given a query image (i.e., Postmortem radiograph), each tooth is matched with the archived teeth in the database (Antemortem radiographs). The experimental results are based on a database of 12 panoramic x-ray images. The experimental results show that the proposed method successfully identifies the individuals based on their dental radiographs.