In last few decades, many attempts have been made on person identification using biometric and non-biometric features. The main objective of person identification is to ensure that the services are accessed only by a legitimate user. But some biological parameters that are used as uni-model biometric fail to provide the guarantee of the presence and aliveness of the person as, voice can be copied through microphone, finger print can be collected on silicon surface, and iris can be copied on contact lenses.
Unimodal biometric systems have to contend with the variety of problems such as noisy data, intra-class variations restricted degrees of freedom, non-universality, spoof attacks and unacceptable error rates. Hence in this project a new approach on person identification system that combines the features of face, finger print and electro-cardiogram which provides an alternative framework of unimodel system leads to multimodal biometrics.
Here, for person identification both biological and physiological parameters was used. Fusion of Two different parameters such as ECG and Blood Vein in Retina is used to identify a legitimate person while avoiding the imposter. Retinal vein authentication which utilizes the histogram equalization for balancing most frequent intensity values with the support of the fast Fourier transform which excels in transforming time domain signals to frequency domain and independent component analysis which is used for ECG feature extraction and finally fusion of these two parameters is passed to support vector machine for the classification Principal Component Analysis method is used for implementation.