Identification Technology Research, Based On The Identity Of The Ecg Signal

Biometrics refers to automatic personal identification on the basis of some physiological characteristics or some behavioral aspects. Because of its many outstanding advantages that other conventional verification methods lack, biometrics has gradually become one of the research focuses of the world. Now, several biometrics that have been used commercially for human identity verification are facial geometry, fingerprints and voice analysis. Unfortunately, face shape can be picked up easily from a photo; people often leave fingerprints on objects which are easy to be stolen by others; and voice can by imitated and so on. So researching the new biometrical identification methods is still one of the focuses in the field of human identity verification.Electrocardiogram (ECG) analysis is not only a very useful diagnostic tool for clinical proposes but also may be a good biometric for human identification. Firstly, it has universality that is to say every person possesses; Secondly, the ECG varies from person to person due to the differences in position, size, anatomy of the heart, age, sex, relative body weight, chest configuration and various other factors. Therefore, it is unique. Thirdly, people's ECG remains unchangeable during a longer time so it is stable. Recently people have begun to use ECG as a means of human identity verification and have got some results.Because the identity verification technology based on ECG is a fairly new method there are still a lot of aspects need to be improved. In order to improve the veracity and the efficiency of identification algorithms we have accomplished following work:(1) Ten people have taken part in our experiment. We collected their ECG signals and gathered 140 ECG data as the original sample.(2) We used wavelet transform to remove the interference of the ECG signal. In the new modified Donoho 's wavelet denoising method we added theoperation of setting the high-frequency detail components of the signal and the low frequency approximation ones to zero. After processing the power line noise, baseline wonder and the high-frequency interference were removed from the ECG signal.(3) We introduced the approach of using the difference-threshold arithmetic in feature extraction. Adopting this method we extracted eight feature parameters: there are P wave amplitude, R wave amplitude, T wave amplitude, P wave duration, PR duration, QRS wave duration, QT duration and ST duration. These parameters would be used during the identity verification.(4) An introduction of BP network and a deduction of BP algorithm were given. Following this a Racial Basic Function network was introduced. In order to identify the center of the network we imposed an improved RBF network construction method .In this arithmetic we combined the clustering method, which is based on comparability threshold and least distance principle with K-means algorithm to identify the center. Then we applied this method to construct the RBF network and used it in the identity verification .Finally, we compared the results of this two methods.The experimental results show that the identification rate of ten people reached 100% with the improved RBF network method. While the identification rate of BP network only 60%. From the result we can see that the RBF network is better than BP network in the veracity and the identification method based on ECG signal is a promising method.