Humans recognize each other according to their various characteristics. For example, a father can recognize his daughter by her face when he meets her and by her voice when she speaks to him. In information technology, biometrics is defined as a method to measure and analyze human body characteristics such as iris, DNA, fingerprints, facial patterns, retina and hand measurements, and so on. For authentication purposes, biometrics is a method of recognizing humans depend on unique physical or behavioral characteristics [1,2]. Not all human traits can be used as a biometric, but it should be characterized by: *Author for correspondence  Universality: all persons must possess the biometric trait.  Distinctiveness: the biometric traits should be unique for each person as DNA.  Permanence: a good biometric trait is invariant or changes slowly over time.  Collectability: a biometric trait should be easily collected quantitatively.In addition, the biometric system should be characterized by: Acceptability: The extent for users to accept the daily using of biometric identifiers.  Performance:according to requirements, accuracy, speed, and robustness.  Circumvention: it should counteract to fraudsters [3].

Biometrics is an interesting study due to the incredible progress in security. Electrocardiogram (ECG) signal analysis is an active research area for diagnoses. Various techniques have been proposed in human identification system based on ECG. This work investigates in ECG as a biometric trait which based on uniqueness represented by physiological and geometrical of ECG signal of person.In this paper, a proposed non-fiducial identification system is presented with comparative study using Radial Basis Functions (RBF) neural network, Back Propagation (BP) neural network and Support Vector Machine (SVM) as classification methods. The Discrete Wavelet Transform method is applied to extract features from the ECG signal. The experimental results show that the proposed scheme achieves high identification rate compared to the existing techniques. Furthermore, the two classifiers RBF and BP are integrated to achieve higher rate of human identification.