|
Descripción
|
|
|---|---|
| Infrared Thermography (IRT) is a widely used technique for the detection of temperature change patterns. In the present study, we are interested in its application to the non-invasive follow-up of temperature changes on emotionally stressed people. Nine subjects performed the arithmetic task of the Trier Social Stress Test (TSST) while IRT videos were acquired. A pipeline was implemented to follow the temperature changes on Regions-of-Interest (ROIs) of the facial IRT videos. The pipeline is divided in four main blocks: (1) Frame selection, (2) image preprocessing (i.e., noise reduction), (3) face segmentation, (4) detection of ROIs (i.e., forehead, nose) and temperature extraction. Block 2 relies in an implementation of Growing Hierarchical Self-Organizing Maps (GHSOM). Faces are segmented (Block 3) with Hybrid Geodesic Region-Based Active Contours (HGRBAC). The detection of the ROIs (Block 4) is performed using the Viola-Jones classification method and finally, the temperature is extracted on the ROIs as a function of the time to enable the identification of stress patterns. Each subject?s performance on the TSST tasks was measured by an Efficiency Ratio, being its value related to the level of concentration of the subject while performing the exercise. | |
|
Internacional
|
Si |
|
JCR del ISI
|
No |
|
Título de la revista
|
IET DIGITAL LIBRARY |
|
ISSN
|
978-1-78561-652-5 |
|
Factor de impacto JCR
|
|
|
Información de impacto
|
|
|
Volumen
|
|
|
DOI
|
10.1049/cp.2017.0148 |
|
Número de revista
|
|
|
Desde la página
|
155 |
|
Hasta la página
|
160 |
|
Mes
|
NOVIEMBRE |
|
Ranking
|
|