Title :
Forehead Thermal Signature Extraction in Lie Detection
Author :
Zhen Zhu ; Tsiamyrtzis, P. ; Pavlidis, I.
Author_Institution :
Univ. of Houston, Houston
Abstract :
Previous work demonstrated that facial thermography can be successful in lie detection. In those studies the development was based on the thermal signature of the periorbital region. In the present paper a new source of psychophysiological information is proposed: the forehead. We found that the corrugator muscle in the forehead is more active than usual, when the individual experiences sustained stress. As a result, more blood flows through the supraorbital vasculature, increasing the cutaneous forehead temperature. In order to monitor the thermal signature of the forehead´s cutaneous tissue, a segmentation method based on active contours has been developed. This creates a virtual forehead probe that can monitor stress levels by measuring thermal radiation over the supraorbital vessels. Thermal videos of 38 subjects under interrogation for a mock crime scenario were used to test the new approach. The results show that the recovered forehead signal, enables 76.3% success rate in deceptive state classification. Thus, the forehead channel shows promise in lie detection.
Keywords :
biomedical optical imaging; biothermics; blood vessels; edge detection; haemodynamics; image classification; image segmentation; infrared imaging; medical image processing; muscle; psychology; active contours; blood flows; corrugator muscle; cutaneous forehead temperature; deceptive state classification; facial thermography; forehead thermal signature extraction; lie detection; mock crime scenario; periorbital region; psychophysiological information; segmentation method; stress level monitoring; supraorbital vasculature; thermal radiation measurement; thermal videos; virtual forehead probe; Active contours; Blood flow; Data mining; Face detection; Forehead; Human factors; Muscles; Psychology; Temperature; Thermal stresses; Algorithms; Body Temperature Regulation; Forehead; Humans; Lie Detection; Models, Biological; Stress, Physiological; Thermography;
Conference_Titel :
Engineering in Medicine and Biology Society, 2007. EMBS 2007. 29th Annual International Conference of the IEEE
Conference_Location :
Lyon
Print_ISBN :
978-1-4244-0787-3
DOI :
10.1109/IEMBS.2007.4352269