BIOMETRIC HUMAN RECOGNITION USING ECG SIGNALS

Abstract: In this research work, ECG of 10 subjects is recorded using the BIOPAC MP 100 system with Acknowledge software. The recorded signal is pre-processed to remove the unwanted high frequency signal, so it is filtered using Low Pass FIR filter with 40 Hz cutoff frequency. Statistical Parameters such as mean, variance, standard deviation etc are used to support the uniqueness property of an ECG signal for each subject. The classifier used in this research work is the Artificial Neural Network (ANN), which is trained with the extracted features. The error in prediction of the ECG of 10 subjects is nearly zero. So the accuracy of the identification system is 100%.

Keywords: Biometric Characteristics, Biometric Traits, Electrocardiogram (ECG), Artificial Neural Network (ANN), Human Recognition.