Authors: K. Sasirekha; K. Thangavel
Addresses: Department of Computer Science, Periyar University, Salem, Tamil Nadu, India ' Department of Computer Science, Periyar University, Salem, Tamil Nadu, India
Abstract: Biometric face classification is an important indexing scheme to reduce face matching time for large volumes of a database. In this paper, a hybridised approach based on rough set theory (RST) and back propagation neural network (BPN) to classify human face is proposed. Local binary pattern (LBP) method is exploited to extract the features from pre-processed face images. The evolutionary optimisation algorithms such as genetic algorithm (GA), particle swarm optimisation (PSO), ant colony optimisation (ACO), hybridisation of ACO and GA (ACO-GA) and hybridisation of PSO and GA (PSO-GA) are investigated for feature selection. Finally, the hybridised rough neural network (RNN) is employed for classification. The experimental results of the proposed RNN is compared in terms of precision, recall, f-measure, accuracy and error rate with Naive Bayes, support vector machine (SVM), radial basis function network (RBFN), conventional BPN, and convolutional neural network (CNN) to conclude the efficacy of the proposed approach.
Keywords: ant colony optimisation; ACO; biometric face; genetic algorithm; GA; gender; particle swarm optimisation; PSO; rough neural network; RNN.
International Journal of Biometrics, 2020 Vol.12 No.2, pp.193 - 217
Received: 06 Sep 2018
Accepted: 10 Sep 2019
Published online: 09 Jun 2020 *