Abstract—Among the various biometric identifications available, fingerprints authentication is the oldest and the most reliable source of authentication. Identification of fingerprints basically relies on minutiae extraction. In most cases the finger print images that are obtained are of poor quality due to various reasons such as scars, dirt, non-uniform ink intensity and skin diversities. So enhancement of image prior to extraction increases the consistency. In this paper a new methodology was proposed for finger print image enhancement where the process begins with ridge orientation using neural network approach that follows dividing the image into white block, black block and grey block, which is called ternarisation, which is a new methodology to deal with uncertainty of minutiae. Then the image is binarised using the concept of pixel aggregation. And finally thinning is applied before proceeding to effective minutiae extraction.
Index Terms—Filling, image enhancement, neural networks, pixel aggregation, ridge orientation, thinning, ternarisation.
M. James Stephen is with Dept of I.T, ANITS, Visakhapatnam, India(e-mail: firstname.lastname@example.org)
P. V. G. D. Prasad Reddy is with Dept. of CS&SE, Andhra Univeristy,Visakhapatnam, India
Cite: M. James Stephen and P. V. G. D. Prasad Reddy, "Enhancing Fingerprint Image through Ridge Orientation with Neural Network Approach and Ternarization for Effective Minutiae Extraction," International Journal of Machine Learning and Computing vol. 2, no. 4, pp. 397-401, 2012.