Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/476648Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Junaidah Mohamed Kassim | |
| dc.contributor.author | Roza Waleed Ali (P84269) | |
| dc.date.accessioned | 2023-10-06T09:23:13Z | - |
| dc.date.available | 2023-10-06T09:23:13Z | - |
| dc.date.issued | 2019-04-23 | |
| dc.identifier.other | ukmvital:123506 | |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/476648 | - |
| dc.description | In the current age, the increase of information and developed techniques growth lead the need for human identification system and information protection. Biometrics system considered as one of the powerful approaches that can be used for human identification system due to its internal feature that is difficult to recreated artificially, stolen and forgotten. However, the biometrics measures are different in terms of the applicability and feasibility of identifying the identification standards such as resistance to aging, power discrimination or resistance to misclassification, and the technology of sensing. In this regard, one of the most potential biometrics measures for identification is a finger vein feature because of a high and unique feature as twin have different vein patterns. In the literature review, there are several algorithms for finger vein identification that differs at recognition stages: pre-processing, region of interest detection, feature extraction, and classification. Other studies have ignored the power of combining different types of features and classifiers in improving the performance of the biometric recognition system. In this study, a novel geometrical feature approach has been developed namely Straight-Line Approximation (SLA) based on database SDUMLA -HMT for finger vein to extend the space of extracting the feature of vein pattern. A set of Extreme Learning Machine (ELM) and Support Vector Machine (SVM) classifiers have been applied. Then, Particle Swarm Optimization (PSO) has been used to improve the result of the (ELM) classifier. The experiment of the optimized extreme learning machine (ELM-PSO) classifier has achieved an accuracy of 88%. Finally, an additional stage namely the combination rules has further enhanced the performance of the finger vein recognition system by combining the results of multiple classifiers excluding ELM-PSO classifier in an overall classifier. The result shows that General Weighted Average Rule (GWAR) and Dempster- Shafer rule (DS) at rank 1 achieved the accuracy of 87% whereas DS and GWAR rules achieved accuracies approximately 93% and 92% at rank 5 in the order.,Master of Computer Science | |
| dc.language.iso | eng | |
| dc.publisher | UKM, Bangi | |
| dc.relation | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
| dc.rights | UKM | |
| dc.subject | Universiti Kebangsaan Malaysia -- Dissertations | |
| dc.subject | Dissertations, Academic -- Malaysia | |
| dc.title | Finger vein recognition using geometrical features and extreme learning machine based on particle swarm optimization (PSO) | |
| dc.type | theses | |
| dc.format.pages | 95 | |
| dc.identifier.barcode | 005474(2021)(PL2) | |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_123506+SOURCE1+SOURCE1.0.PDF Restricted Access | 1.67 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.