This work presents a review of different indexing techniques designed to enhance the speed and efficiency of searches over large biometric databases. The coverage includes an extended Delaunay triangulation-based approach for fingerprint biometrics, involving a classification based on the type of minutiae at the vertices of each triangle. This classification is demonstrated to provide improved partitioning of the database, leading to a significant decrease in the number of potential matches during identification. This discussion is then followed by a description of a second indexing technique, which sorts biometric images based on match scores calculated against a set of pre-selected sample images, resulting in a rapid search regardless of the size of the database. The text also examines a novel clustering-based approach to indexing with decision-level fusion, using an adaptive clustering algorithm to compute a set of clusters represented by a ‘leader’ image, and then determining the index code from the set of leaders. This is shown to improve identification performance while using minimal resources.
This work presents a review of different indexing techniques designed to enhance the speed and efficiency of searches over large biometric databases. The coverage includes an extended Delaunay triangulation-based approach for fingerprint biometrics, involving a classification based on the type of minutiae at the vertices of each triangle. This classification is demonstrated to provide improved partitioning of the database, leading to a significant decrease in the number of potential matches during identification. This discussion is then followed by a description of a second indexing technique, which sorts biometric images based on match scores calculated against a set of pre-selected sample images, resulting in a rapid search regardless of the size of the database. The text also examines a novel clustering-based approach to indexing with decision-level fusion, using an adaptive clustering algorithm to compute a set of clusters represented by a ‘leader’ image, and then determining the index code from the set of leaders. This is shown to improve identification performance while using minimal resources.
This will help us customize your experience to showcase the most relevant content to your age group
Please select from below
Login
Not registered?
Sign up
Already registered?
Success – Your message will goes here
We'd love to hear from you!
Thank you for visiting our website. Would you like to provide feedback on how we could improve your experience?
This site does not use any third party cookies with one exception — it uses cookies from Google to deliver its services and to analyze traffic.Learn More.