In this article, both the basic principles and the experimental methods of identity recognition technology based on heart sounds are addressed. First, the characteristics of heart sounds and the feasibility of heart sounds as a biometric were analyzed. A synthetic model of heart sounds was then developed based on a family of wavelets; Finally, the characteristic parameters of heart sounds were extracted by using the heart sounds linear band frequency cepstra (HS-LBFC) with a specified configuration, and the similarity distance was adopted for heart sound identification. To highlight the difference in two heart sound signals between the time and frequency domains, a construction method of the heart sound wavelet, a calculation method of the parameters of a synthetic model, selection of the characteristic parameters of heart sounds and the corresponding data processing technology were examined. The experimental results showed that the proposed method exhibited excellent performance and practicability.Keywords biometric, wavelet family of heart sounds, synthetic model of heart sounds, heart sounds linear band frequency cepstra (HS-LBFC)