Study of drowsiness from simple physiological signals testing: A signal processing perspective
| Autor: | Charat Thothong, Pornkanok Sukaimod, Pinkaew Khongsabai, Noppawit Aeimpreeda, Direk Sueaseenak |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
0209 industrial biotechnology
Signal processing Computer science Speech recognition media_common.quotation_subject Perspective (graphical) 02 engineering and technology Signal 020901 industrial engineering & automation Feeling Filter (video) QUIET 0202 electrical engineering electronic engineering information engineering Graph (abstract data type) 020201 artificial intelligence & image processing Sleep (system call) media_common |
| Zdroj: | ICAIIC |
| DOI: | 10.1109/icaiic48513.2020.9065035 |
| Popis: | This paper reports a study of simple physiological signals that are electrodermal activity, pulse rate, and head nodding in drowsiness state and normal state. For the experiment protocol, the subjects were sat in a dim and quiet place and physiological signals were collected for 15 minutes by using Biopac MP36. A questionnaire was used to assess subject's feeling before and after the experiment. To classify the state of drowsiness, we have followed the criteria of previous study. Low-pass filter 2–5 Hz was used when pulse rate and electrodermal activity were recorded. In first row of a graph is an electrodermal activity that does not pass filter this is will define about the filter Low-pass band in how many hertz we use to filter in this electrodermal activity graph. All of data in pulse rate and electrodermal activity that we collected is from Biopac MP36. Then we collected head nodding in degree from Arduino mega with acceleration sensor module and example of data is in Figure 3. Thus, this paper uses all collected data and questionnaires to be observed and studied all signal and trend about the change rate in drowsiness state and normal state of people. All in all, we found that Electrodermal activity has their own platform represent to drowsiness stage. Electrodermal activity signals decreasing continually similar graph of cos θ and stop going down if human is sleeping peacefully. For pulse rate signals, the graph always fluctuated during drowsiness and normal stage. Head nodding cannot define a differentiate between normal stage and drowsiness stage |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|