Environment aware chemical detection & classification

Each chemical agent especially, explosive chemical agent exhibits a distinct infra red (IR) absorption spectrum or signature. To identify them uniquely, we need high resolution spectrometers in the absence of smart algorithms. Such spectrometers´ optical components are bulky and expensive. The cost of a spectrometer can be traded with resolution. In such a situation, the detection and classification algorithm should be smart. The smartness is coming from the algorithms that extract unique features and recognize the patterns in the features. In this paper, we describe these algorithms and provide the classification results obtained by using IR absorption spectra of different explosive chemical agents. Our preliminary results of classification of 14 explosives indicate that we can obtain 100 % classification accuracy.