Heart Sounds in Identity Recognition

In this thesis the applicability of advanced signal processing techniques to the status distinguishes using heart sound is first demonstrated. Two advanced signal processing techniques namely: algorithm of independent sub-band functions and the heart sound cycle frequency spectrogram. After analyzing the basic characteristics of heart sound signals, the characteristic parameters of heart sound human are extracted by algorithm of independent sub-band functions; identity recognition of single channel heart sound signal was achieved by setting up "S1-S2-CFS" recognition model and using pattern matching method of similarity coefficient. The final experiment results demonstrate that the new technology is better.