| Title |
Personal Authentication Method using Segment PPG signals |
| Authors |
이선우(Sun-Woo Lee) ; 우덕균(Duk-Kyun Woo) ; 마평수(Pyeong-Soo Mah) |
| DOI |
https://doi.org/10.5573/ieie.2019.56.6.45 |
| Keywords |
PPG ; Biometric ; Wearable device ; Random Forest ; |
| Abstract |
A study regarding personal authentication based on segmented PPG has been conducted using Hybrid KNN Random Forest algorithm. It is important to divide the PPG data into segments to authenticate the person. When PPG data is divided into segments, each segment is set so that Peak, Qnet, and Dicrotic notch data can be contained in one segment. For the consistency of the data between each segment, the data were preprocessed using normalization and interpolation. Using the preprocessed PPG data, we used a instantaneous frequency based on symmertric higher order differential energy operator and poincare plot using the ratio of SD1 and SD2. Then, the extracted features are used as input variables for machine learning technique such as KNN, Random Forest, Hybrid KNN Random Forest. The accuracy of the algorithms is 78%, 95%, 96% respectively. |