Identification of people from armpit odor region using networked electronic nose

Abstract

Homeland security basically needs devices that can detect, track, and identify a terrorist from a distance. Body odor recognition offers an opportunity to confirm a person's identity based on one's unique odor pattern. In this paper, we have reported how to invent a networked electronic nose (E-nose), which can detect and recognize human odors from the armpit region. An array of metal oxide sensors was used to detect human odor. Principal component analysis (PCA) was employed to perform pattern recognition and discrimination. The method for correction of the sensor drift has been proposed. The results show that the networked E-nose has a capability to detect human body odor and can create the unique smell print of each person. Based on PCA with 95% confidence ellipse, this E-nose can identify a person from four persons by detecting odor from armpit region. © 2011 IEEE.