Photoplethysmographic authentication in long-term scenarios: a preliminary assessment

The photoplethysmogram (PPG) is a biomedical signal that estimates the skin blood flow by means of infrared light. It is normally used to measure the oxygen saturation and the heart rate. During the last few years, some studies have been published to asses the PPG potential for its use in biometric authentication systems, but results are inconclusive. In this paper we propose a new approach for PPG biometric identification based on the Karhunen-Loève transform for the feature extraction and on the Euclidean distance for the classification. Using a public PPG database and a small local database, we have evaluated the possibility of using the PPG for authentication in the long term run. The new method obtains an equal error rate (EER, value where the proportion of false acceptances is equal to the proportion of false rejections) value of 4.8% in the best case, degrading to 20% in the long-term run, which suggest a performace degradation with time.