Please use this identifier to cite or link to this item: http://repository.futminna.edu.ng:8080/jspui/handle/123456789/28581
Full metadata record
DC FieldValueLanguage
dc.contributor.authorAbdullahi, Muhammad Bashir-
dc.contributor.authorIdris, Fati-
dc.contributor.authorAlhaji Mohammed, Adamu-
dc.date.accessioned2024-05-20T12:57:24Z-
dc.date.available2024-05-20T12:57:24Z-
dc.date.issued2016-12-
dc.identifier.citationM. B. Abdullahi, F. Idris and A. A. Mohammed, "Performance analysis of particle swarm optimization algorithm-based parameter tuning for fingerprint image enhancement," 2016 Future Technologies Conference (FTC), San Francisco, CA, USA, 2016, pp. 528-536, doi: 10.1109/FTC.2016.7821658.en_US
dc.identifier.isbn978-1-5090-4171-8-
dc.identifier.urihttp://repository.futminna.edu.ng:8080/jspui/handle/123456789/28581-
dc.description.abstractExisting algorithms designed for Fingerprint Image Enhancement either lack the ability to enhance poor quality image or are computationally expensive. Evolutionary algorithms are often used to enhance images. Particle Swarm Optimization (PSO) is one of the most progressive algorithms but has parameters, which are not properly tuned to reduce the number of iterations. In this paper, PSO parameters; inertia weight (w) and acceleration constants (c1 and c2) were fine-tuned. PSO-based parameterized transformation function, which incorporates both the global and local information of an image was developed to maximize the information content of the fingerprint image. In the transformation function, a threshold of 0.99 was set to control the contrast effect of the enhanced image. The intensity values of pixels that are less than the threshold were transformed. The image quality was evaluated using an Objective Function in term of Number of Edges, Sum of Edge intensities and the exponential of the entropy. The commonly-well-known database FVC-2004 is used in this study. It was observed from the experiments that the best PSO parameters set used for successful convergence of the PSO Algorithm were w ∈ [0.7, 0.75] and (c1, c2) ∈ [1.2, 1.3]. Therefore, any set of values used outside these ranges would result to local minimum convergence and increase the computational effort by searching in unwanted areas.en_US
dc.language.isoenen_US
dc.publisherIEEEen_US
dc.subjectparticle swarm optimizationen_US
dc.subjectfingerprint image enhancementen_US
dc.subjectparameter tuningen_US
dc.subjecttransformation functionen_US
dc.titlePerformance Analysis of Particle Swarm Optimization Algorithm-Based Parameter Tuning for Fingerprint Image Enhancementen_US
dc.typeArticleen_US
Appears in Collections:Computer Science



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.