Please enable JavaScript.
Coggle requires JavaScript to display documents.
Fingerprint Recognition in Forensic Scenarios - Coggle Diagram
Fingerprint Recognition in Forensic Scenarios
Abstract
Fingerprints=unique biometric key that allow for identification
A system that can match the details of different images in large databases are still a problem
Collection at crime scenes are manually processed
Proposed methodology: use of a Gabor filter for pre-processes, extraction of minutiae using the crossing numbers method, and replacing two or more very close minutiae with average minutiae, creation of a model that represents each minutiae through polygons, and the individual search of a match for each minutia in different images using metrics and relative errors.
Differs: others validate the entire fingerprint model while they want to validate each minutia using n-vertex polygons
Key words
Fingerprints
Biometrics
Polygons
Minutiae
Forensic analysis
Intro
Technological advances processes enable capture, storage and comparison
Manual collection takes a long time that can be used to do something else
A lot of time is spent on comparison when a computer can do it in seconds
Images in the field can be of poor quality
Methodology propose that can compare two fingerprints on a mobile device and validates each minutia individually
New process to validate extracted minutiae using convex hull and employing n-side polygons
Background
Fingerprint features
Bifurcation and termination
Cross numbers method
Related work
Databases: FVC is most common