Affiliations: [a] Department of Electronics and Communication, Bundelkhand Institute of Engineering and Technology, Jhansi, India | [b] NSIT, New Delhi, India
Correspondence:
[*]
Corresponding author: Sharad Pratap Singh, Department of Electronics and Communication, Bundelkhand Institute of Engineering and Technology, Jhansi, India. %****␣kes-25-kes210068_temp.tex␣Line␣25␣**** E-mail: [email protected].
Abstract: Fingerprint matching is based on the number of minute matches between two fingerprints. Implementation mainly includes image enhancement, segmentation, orientation histogram, etc., extraction (completeness) and corresponding minutiae. Finally, a matching score is generated that indicates whether two fingerprints coincide with the help of coding with MATLAB to find the matching score and simulation of Artificial Neural Network extending the feedback of the network. Using the artificial neural network tool, an important advantage is the similarity index between the sample data or unknown data. A neural network is a massively parallel distributed processor consisting of simple processing units that have a natural property to store knowledge and computer experiences are available for use. A fingerprint comparison essentially consists of two fingerprints to generate a fingerprint match score the match score is used to determine whether the two impressions they are of the same finger. The decision is made this study shows the comparison of normal and altered fingerprints using MATLAB coding and data used to study in the self-generated data using biometric scanner also the open source data available on the web is used for finding out matching score or similarity index, The study shows that there is hardly any matching between normal and altered fingerprints of the same person.
Keywords: Artificial neural network, feedforward back propagation, fingerprints, minutiae, biometrics
