A Challenge to Analyze and Detect Altered Human Fingerprints

Abstract

The deployment of Automated Fingerprint Identification Systems (AFIS) in law enforcement and border control applications has escalated the need for ensuring that these systems are not compromised. Several problems related to fingerprint system security have been investigated carefully;, the problem of fingerprint alteration has received very scarce attention (1). Fingerprint alteration refers to the deliberate alteration of the fingerprint pattern by a person for masking his/her identity. Several cases of fingerprint alteration have been reported previously. The main contributions of this paper is: 1) Analysis of the altered fingerprints 2) classifying the alterations into three major categories: Obliteration, distortion and imitation and suggesting possible countermeasures 3) Demonstrating by using an example of database, where based on matching score the person of interest is branded as guilty or not and then based on decrease in value of matching score(4)(1) if below threshold value we can declare that the fingerprint is altered or not then combine fingerprint Image with person's UID(Unique Identity) comparing with that fingerprint with existing fingerprints showing if there is any match in the form of percentage and display his/her details with criminal record, 4) developing the above system by a technique to automatically detect altered fingerprints based on segmentation using Morphological operations, minutia marking with special considering the triple branch counting(11), minutia unification by decomposing a branch into three terminations(11)(12), and matching in the unified x-y coordinate system orientation estimation(13), minutiae matching after a two-step transformation.