The process of fingerprint thinning creates a 1-pixel wide pattern of the fingerprint ridges while retaining the basic fingerprint structure. Several algorithms have been devised [1 to 7] to extract the thinned pattern of a fingerprint image; however, such algorithms produce different results with different positions and rotations of the same fingerprint image, which results in inefficient minutia extraction. In this paper, a new method of fingerprint thinning is proposed: Minutia Analysis using Rotation Invariant Algorithm (MARIA), which thins a fingerprint image regardless of the fingerprint position and rotation and then extracts minutia from the thinned pattern. The fingerprint image is binarized to convert it into a 0-1 pattern. Morphological operations are applied to remove islands and lake- like portions from the image. If-then rules governing a 3x3 mask that is to be convoluted throughout the image are applied to thin the transformed reduced image in a rotation-invariant manner. Post processing is performed on the thinned image to remove any spurious minutia points. Finally, genuine minutia points and their directions are extracted from the thinned fingerprint image. The experimental results indicate that MARIA successfully extracts genuine minutiae from low-quality fingerprint images, and greatly facilitates the minutia matching procedure.
Fingerprint; Thinning; Minutiae; Dilation, Back-propagation Neural Network; Minutia matching.
To return to the Volume/Number webpage, click here.
THE INTERNATIONAL JOURNAL OF FORENSIC COMPUTER SCIENCE - IJoFCS
Volume 5, Number 1, pp 8-15, DOI: 10.5769/J201001001 or http://dx.doi.org/10.5769/J201001001
MARIA: Minutiae Analysis using Rotation Invariant Algorithm
By Avinash Pokhriyal and Sushma Lehri
To download this paper, click here