Efficient Biometric Security System Using Intra-Class Finger-Knuckle Pose Variation Assessment

Abstract

Finger Knuckle Print is a finger based biometric system. It is a model which is the most widespread biometric authentication for an individual. The performance of Finger Knuckle Print recognition algorithms mostly depends on different pose variations. Due to deformations on the fingerknuckle samples, the result of the pose variations leads to false accept rate and it reduces the robustness level. In this paper, an efficient method is introduce to solve the intra-class finger-knuckle pose variations, called the Kernel Intra-Class Finger-Knuckle Pose Density Assessment (‘K’ Intra-Class FKPDA) to improve the robustness of the biometric system. ‘K’ Intra-Class FKPDA method map the user’s fingerknuckle prints samples in high dimensional feature space using the kernel process. Markov Graphic Model Mapped finger-knuckle print samples on high dimensional feature space to calculate the density level and make out the intra-class pose variations. Bessel Thomson Finger-Knuckle Pose Variation filtering process identified intra-class pose variation and it is removed in ‘K’ Intra-Class FKPDA method. To reduce the delay time ‘K’ Intra-Class FKPDA method applies a linear filtering model. Finally, using ‘K’ Intra-Class FKPDA removed intra-class pose variations and finger-knuckle print is documented with increased robustness level. PolyU FKP Database evaluated Experimental results on 165 images show that the new method can improve the robustness factor appreciably compared with typical Finger Knuckle Print based recognition approaches reducing the filtering delay factor. Experiment is conducted on false acceptance rate, recognition rate, and filtering delay time.

Authors and Affiliations

Mr. J. Stanly Jayaprakash , Dr. S. Arumugam

Keywords

Related Articles

CRM Based On Naive-Bayesian Classification

CRM is a combination of several components. The companies can process the customer information before beginning the process. Companies can acquire about their customers through their internal customer information or they...

Handwritten Character Recognition for major Indian Scripts:A Survey

Character Recognition is an active field of research today. It comprises of Pattern Recognition and Image Processing. Character Recognition has gained enormous attention due to its application in various fields. Characte...

Database Privacy- Issues and Solutions

Data Mining, fourth and analytical step of Knowledge discovery in database process is a process of discovering new and interesting patterns in the large datasets. For example, data miner can derive different patterns bas...

Performance Evaluation of BST Multicasting Network over ICMP Ping Flood for DDoS

The Paper evaluates the performance of Bi-directional Shared Tree (BST) multicasting network with two sources and four receivers, attacked by five attackers. These attackers attack the source 1 in network by ICMP Ping Fl...

Analyze the impact of Transmission rate on the Performance of AODV and DSR Protocols in MANETs under Responsive and Non-responsive Traffic.

The massive boom in the wireless technology has led to heavy utilization. Due to the heavy utilization and shared nature of resources causes QoS problems in ad hoc networks. Providing QoS is a severe problem in mobile ad...

Download PDF file
  • EP ID EP132147
  • DOI -
  • Views 102
  • Downloads 0

How To Cite

Mr. J. Stanly Jayaprakash, Dr. S. Arumugam (2014). Efficient Biometric Security System Using Intra-Class Finger-Knuckle Pose Variation Assessment. International Journal of Computer Science & Engineering Technology, 5(12), 1114-1119. https://europub.co.uk/articles/-A-132147