Multimodal Biometric Authentication Using Back Propagation Artificial Neural Network

Abstract

Recently, the biometric-based identifications are widely adopted for personnel identification. However, the unimodal recognition systems currently suffer from noisy data, biometric sensor data quality, spoofing attacks, unacceptable error rates and lack of distinctiveness of the biometric trait. These issues can be undertaken via multi-modal biometrics authentication system. This paper proposes a multi-modal framework to capture human skeletal features and facial features using imaging techniques. This modelling technique captures human joints with 3D depth data, to improve the system efficiency. This Biometric technique is used to recognize the known/unknown image from the trained images within short span of time. The proposed research has subdivided into three-folds. First, preprocessing the image, using the MinMax method which reduces the noise in data and enhance the image quality. Second, image features are extracted using Genetic method in combination with cuckoo search algorithm which is considered to be best optimization technique for image analysis. Third, the human recognition is attained via the Artificial Neural Network (ANN) which relies on back-propagation algorithm. The performance of the proposed method is evaluated based on the measure of different metric named as, true positive and negative rate, false positive and rejection rate, accuracy. The experimental results show that this new framework dramatically improves the system performance.

A Computational Approach for Score-Level Fusion Decision-Making of Multi-biometric Recognition System Using Ant Colony Optimisation

“…They finished the match with a combined score of 101.0414%. Priya and Mukesh [15] researched human skeletal and facial traits to develop a biometric system for human identification. They first preprocessed the photos, extracted their characteristics, and then used ANNs to classify the images.…”

Odour Based Human Identification and Classification using Neural Networks

“…Efficient and accurate results obtained for fingerprint detection and face recognition using artificial neural net [2,8,11,12,15]. Palm print and biometric hands detection employing back-propagation neural net with Levenberg-Marquardt training algorithm used in the classification of Palm print Biometrics.…”

“…Many personal characteristics, imaging methods, and body parts have been utilized as biometric systems such as teeth, ears, fingers, voices, hands, feet, eyes, gaits, veins, typing styles and signatures. Each biometric takes its own limitations and strength, so that each biometric is employed in applications of Identification (authentication) [2].…”