Search

Since the late 1800s, fingerprints have been used as a technique of identifying persons. Biometrics technology, which employs physical or behavioral features to identify people, is gaining popularity as a reliable and effective method of personal verification that aids in the identification of a real person. As a result, utilizing MATLAB, this study provides an effective method for automating fingerprint recognition for the verification and authentication of specific people. Human fingerprints contain minute characteristics known as minutiae that can be utilized as identity marks for fingerprint verification. Fingerprint identification is amongst the most effective and popular biometrics, due to their distinctiveness and stability over time. But due to developments in computing power, they have only recently become automated (i.e., biometric). Fingerprint recognition software comes in a number of types and sizes, and it can be utilized for a wide range of purposes. One of the most stimulating pattern recognition problems is fingerprint recognition. The Fingerprint Recognition system consists of three steps. Pre-processing increases binarization and thins the fingerprint image in the first step. The second process is the Feature Extraction Stage, involving the employment of minutiae extractor techniques for extracting ridge termination and ridge bifurcation from the thinned image. The third process is matching (identification and verification), which requires using the minutiae matcher process, that uses distance and similarity metrics to match two minutiae points. Fingerprint images from the FVC2002 database are used to evaluate the algorithm's accuracy and reliability. The testing database for this research, on the other hand, is made up of fingerprints gathered from an online database. [ABSTRACT FROM AUTHOR]

This paper presents a modified AMRI conjugate gradient (CG) technique based on exact line search. The novel approach is use for solving unconstrained optimization problems. The performance of the proposed algorithm based on number of iteration and CPU was recorded. Under some suitable conditions the convergence of the proposed method is discussed. Finally, new method was extended to estimate unemployment rate in Nigeria. [ABSTRACT FROM AUTHOR]