Multimodal Biometric Identification System

Biometrics provides greater security and usability than conventional identification methods. In order to render an effective biometric system this book provides an improved multi-model biometric recognition based on iris and fingerprints. The work is established on three modules which are pre-processing module, feature extraction module and recognition module. At first, pre-processing is performed through histogram equalization on input image to enhance its quality. Then, the extraction of the feature set is performed which include modified Local Binary Pattern (MLBP), GLCM with orientation transformation, and DWT features. Consequently, the optimum function is found with the Rider Optimization Algorithm(ROA). Fusion of optimized fingerprint and iris features is carried out with the help of Fish Swarm optimization algorithm. Further, Improved Multi Kernel Support vector machine (IMKSVM) and Deep Neural Network (DNN) algorithms are used for authentication process. In IMKSVM, several kernels are integrated to give shape to a hybrid kernel. Alternatively, DNN is a multi-layered artificial neural network which finds the right mathematical transformation to turn input into output.