A Study on the Possibility of Measuring the Non-contact Galvanic Skin Response Based on Near-Infrared Imaging

Galvanic skin response (GSR) is the change in the electrical conduction of the skin caused by an emotional response to a stimulus. This bio-signal appears to be the action of the autonomic nervous system as a change in sweat gland activity and reflects the intensity of emotion. There have been various technologies for detecting autonomic nerve reactions to bio-signals in image processing; research on non-contact detection of these technologies has become amajor issue. However, research onGSRfor detecting signals in a non-contactmethod is lacking compared to other bio-signals. In this paper, we propose a non-contact GSR detectionmethod from the human face region using a near-infrared camera. The proposed method used a 20-min visual stimulus video that caused tension or excitement among the participants and set up an isolated experimental environment to minimize emotion induction caused by external factors. Afterward, a GSR sensor was attached to the inside of the participant's finger, and a visual stimulus video was shown. The participant's face region was shot using a near-infrared camera, which also stored GSR sensor data and infrared face images that were input during the experiment. The face region was detected from the acquired face images, and the region of interest inside the face was compared with the data obtained through the GSR sensor. Finally, the correlation between the image intensities inside the designated region of interest and the data from the GSR sensor were analyzed.