Akademik Veri Yönetim Sistemi
25th Signal Processing and Communications Applications Conference (SIU), Antalya, Türkiye, 15 - 18 Mayıs 2017
A lot of information can be attained with analysing biological signals which are electroencephalogram, electrocardiogram, electromyogram, magnetoencephalogram and photoplethysmography (PPG). This information is utilizez for in both diagnosis and criminal research. Photoplethysmography is a painless, simple and inexpensive optical technique that can be used to detect blood volume changes in microvascular tissue bed. In this technique, the measurement is taken from the under part of the right forefinger. In this study, photoplethysmography data of 66 participants from Karadeniz Technical University Electrical and Electronics Engineering, 46 of whom were male, were made to identify the smoking habits, gender and starvation status by extracting features from these biological signal. In this study photoplethysmography signals are represented both by the skewness, kurtosis, maximum, minimum which is statistical value and continuous wavelet transform coefficients features. The extracted features are classified by the k-nearest neighbourhood method, which is fast and easy to apply. The detection of smoking, gender and starvation status were classified with an accuracy of 73.7%, 72.8% and 65.8%, respectively. The results have shown that people which smoking habits, gender, starvation can detect from PPG signal.