Heart Sound Identification System Based On Android Smartphone

With the rapid development of information technology, a higher requirement on personal information security has been put forward. The traditional identification technology is unable to meet the demand of people for highly accurate identification methods. Biometric identification technology, as an approach of using body physiological or behavioral characteristics to realize identification with higher safety and reliability, has gradually become the international research hotspot, and played an important role in different fields. The common biometric identification technologies include iris, fingerprint, retina, speech, face and DNA recognition. Despite of these identity recognition technologies being utilized in the practical application successfully, some accompanying security challenge make exploring new biometric characteristics identification methods as one of the new trends of biometric identification technology.This paper utilize s heart sound of human body as new biometric characteristics and proposes a new heart sound identification system based on Android operating system. Based on some relevant researches of key technologies, such as characteristic extraction of heart sound and pattern matching, the heart sound hardware acquisition device is designed and related Android application program for heart sound identification is developed in Android smart phone. The researches of this paper are mainly as follows:(1) The principle of heart sound identification is introduced, and the history of heart sound identification and current related researches are reviewed. The Android system architecture, characteristic, application components and the construction of development environment are introduced.(2) The heart sound acquisition device, which mainly consists of heart sound sensor, filter circuit, dsPIC main control circuit, gain control circuit, HC-06Bluetooth transmission module and NFC tag, is designed to collect heart sound signal. Heart sound is acquired by heart sound sensor and pre-processed by filter circuit, and then A/D conversion of heart sound is completed by dsPIC main control circuit, finally the wireless transmission of heart sound is realized by NFC tag and Bluetooth transmission module. The main function of NFC tag is to simplify the process of Bluetooth pairing, establishing the connection between Android smart phone and heart sound acquisition device quickly to ensure operating experience of users.(3) The identification algorithm based on MFCC and improved VQ is proposed. Firstly, the pre-processing method of heart sound segmentation based on high-order Shannon entropy is proposed, and secondly, the MFCC feature is extracted according to segmented heart sound. Thirdly, the principle of VQ model is discussed and LBG algorithm for codebook training is introduced. Finally, the weighted distortion measure VQ based on variance and standard deviation is proposed.(4) The application program based on Android operating system is developed, which mainly consists of four functional modules:Bluetooth connection, user registration, user identification and system management. NFC technology is applied in Bluetooth connection module to simplify the Bluetooth connection. In user registration module, the users’basic information is registered and heart sound is acquired and then stored in Android smart phone. The heart sound feature is extracted by MFCC algorithm and the codebook is trained by LBG algorithm. The user identification module is used for human identification, and the main function of system management module is to manage the information of registered users.In this paper, a heart sound identification system based on Android is proposed, which can provide a reliable, secure and portable device for biometric characteristic identification field.