Development Of Embedded Identification System Based On The Analysis Of Heart Sound

In modern society，with the rapid development of transport, communication and networktechnologies, also as the increasing scope of human activities, the difficulty and importance ofidentification are becoming more and more prominent. We put forward higher request aboutaccuracy, security and availability of identification system. Biometric identification technology isthe way which makes use of the human physiological characteristic and behavior of theinherent characteristics through computer, optical, acoustic, biological sensors and other hightechnology. Though traditional biometric technology such as fingerprint, iris, palm print and voicehave being successful in practical application, they are facing the security problem that theidentities fake easily. We need new techniques to overcome the insecurity and unreliability. As aninternal physiological signal of human body, heart sound is expected to become a new identificationmethod because of its unique physical characteristics.In this paper, we develop an embedded identification system by the analysis of heart sound.Based on the research in the heart sound signal preprocessing, feature extraction andmatching model，we develop the hardware and software platform of the heart sound identificationsystem, and design the identification software. Eventually，we realize the embedded heart soundidentification system. The main contents are as follows:1. Research the heart sound identification theory and provide the theoretical basis for featureextraction. Review the history of heart sound identification technology and the status quo at homeand abroad.2. Combined with VQ (Vector Quantization), we put forward the WPCC (Wavelet PackageCepstrum Coefficients) method in LabVIEW to get the sole and stable heart sound code which canbe used as the characteristic for identification.3. Design the embedded heart sound identification system equipment, including the uppercomputer and lower computer. The lower computer includes the heart sound pickup head, activefilter amplifier, main amplifier, dsPIC control unit, HC-06 Bluetooth transmission circuit andpower circuit, the host computer includes the HC-06 signal reception circuit, industrialmotherboard and LCD screen. The Lower machine acquires the heart sound signals. Through thefilter-amplifier, it will be fed into the AD conversion to generate the sampling signal which will betransmitted by the Bluetooth. Taking charge of the human-computer interaction and recognitionalgorithm processing, the host computer receives the sampling signal and achieves theeffective recognition of heart sound. 4. Taking the advantage of Microsoft Access 2007, we develop the heart sound identificationsoftware in LabVIEW, build the heart sound database and realize the function of data acquisition,user registration, user identification and system set. Finally, the seamless migration of the embeddedplatform is accomplished through the setup which is created by the LabVIEW engine.5. Design the Windows Embedded Standard operating system that not only realizes thefunction of operating system customizing, drive customizing and shell customizing, but also carriesthe software developed in LabVIEW to achieve the goal of heart sound identification easily andflexibly.This study provides the biometric identification equipment with high accuracy and wellanti-counterfeiting ability which will be an ideal solution for the various identification problems andsecurity challenges.