Personal Shirt Odor Classification Using an Electronic Nose

This paper presents the classify personal shirt odor by electronic nose. Because body odor of a particular person varies, the researcher has picked up 4 volunteers to do exercise for 2 hours. After that bring their shirts to evaluate volatile organic compounds (VOCs) from the dirt on that shirts. The resistance of gas sensor array in the electronic nose will be changed depend on the amount VOCs when various kinds of gas sensor are combined together, the result will vary according to the property of each particular gas sensor. Then evaluate highest signal changing value of gas sensor and analyze by Principal Component Analysis (PCA). This technique is to identify the influence of the gas sensor array on the volunteer's shirts odor. From the experiment, we found that electronic nose can identify the odor of each shirt well, the data has been separated to 4 groups after analyze by PCA technique. It can extract data relation and percentage accumulation of PCA1 and PCA2 are 82.39.