Font Size: a A A

Heart Sound Identification System Based On Android

Posted on:2015-03-25Degree:MasterType:Thesis
Country:ChinaCandidate:W LiuFull Text:PDF
GTID:2298330467477025Subject:Electronic and communication engineering
Abstract/Summary:PDF Full Text Request
The traditional identification technology is mainly characterized by personal password,it isdifficult to meet the high security requirements of network information. Compared with traditionalidentification technologies, biometric identification technology is safer, more reliable, moreconvenient and more accurate. Heart sound signal is a normal physiological signals, which is unique,universal and stability. Therefore, the heart sound signal is difficult to be copied, which makes heartsounds into a high security biometric. This paper focuses on the identification of heart sounds basedon DTW, uses MFCC parameters for the identification of heart sounds and designs the heart soundidentification system based on the Android platform.This paper introduces the background, significance and status of domestic and internationalresearch of heart soundidentification. Then it introduces the Android operating system whichincludes the origin, architecture, applicationscomponentsand software development environmentofAndroid.Thenit introduces the basic theory of heart sound which includes the composition and themechanism, time domain and frequency domain characteristics, authentication framework andperformance metrics of heart soundidentification.Then it introduces pretreatment, endpointdetection,extraction and pattern matching of heart sound heart. Finallyit introducesdetails of thesystem requirement, interface and code ofrecognition algorithm.The first step of heart sound identification is pretreatment. The second step is use dual-thresholdendpoint detection methods to accurately extract the start and end of heart sound signal.The thirdstep isto use MFCC algorithm to extract characteristics parameters of heart sound, then use DTWalgorithm to pattern matching. The last step is to choose the heart sound which has the minimumEuclidean distance as the result.The system interface is simple and easy to operate and has a fast recognition speed. What’smore, it meet the design requirements and achieve a Android machine running. It achieve thedesired effect and has some pratical value.
Keywords/Search Tags:Authentication, Heart sound recognition, Android, DTW, MFCC
PDF Full Text Request
Related items