Multimodal biometric recognition using fused handcrafted features

Multimodal biometric systems have been widely implemented in a variety of real-world scenarios due to their ability to overcome limitations associated with unimodal biometric systems. This paper is focused on the combination of the face, ear and gait in a unified multimodal biometric identification system using handcrafted features. These approaches provide robust and discriminative features to solve the biometric problem. In this research, speed up robust features and histogram of oriented gradients approaches have been used to extract features from face, ear and gait. The extracted features are optimized using genetic algorithm and classified using Levenberg-Marquardt backpropagation neural network. The system performance is evaluated on constrained and unconstrained dataset conditions.