Implementation of ICA based Score level Fusion of Iris and Ear Biometrics

Biometric systems based on multimodality verifies the uniqueness of an individual that depends on behavioral biometric traits such as signature and voice or physiological traits such as face and fingerprint. The objective of the biometric systems is to narrow down the irresponsibleness by finding the correct match and in addition it will increase the safety level in contrast to the arrangements which are made using one biometric attribute. In this work, an advanced biometric system, which depends on multimodality, is developed, i.e., using ear and iris. Firstly, Ear and Iris images have been uploaded for recognition. Secondly, independent component analysis (ICA) will be used for feature extraction. Finally, genetic algorithm has been used for the optimization purpose. Later, the ear and iris trait area unit is combined with hamming distance and the results are confirmed for authenticity. For the realization of the results of the intended method, false rejection rate (FRR) and false acceptance rate (FAR) are used and it is implemented in MATLAB environment. Both quantitative and qualitative results show significant improvement in the technique using Hough man circular transform (HCT) and ICA over principle component analysis (PCS) with ICA.