The principal goal of sensors for the detection of explosives is to establish the identity of the interrogated target as a key to threat assessment and decision making. Despite the fact that both Raman spectroscopy and laser-induced breakdown spectroscopy (LIBS) have shown their capability in standoff detection of explosives, such techniques are not exempt from certain limitations, in terms of sensitivity and selectivity, to carry out this purpose when they are used individually. For this reason, the idea for the fusion of data reported by these orthogonal techniques, Raman and LIBS, has been around for a while. The present manuscript proposes an approach for the combination of the spectral outputs of Raman and LIBS sensors (data fusion strategy) in order to obtain knowledge about the identity of compounds better than that achieved when each technique acts alone. After simple mathematical treatment, a new pattern of identification (two-dimensional, 2D, image) of several compounds (explosives, confusants, and supports) was generated from the assembly of their Raman and LIBS spectra. The efficiency of this strategy was evaluated by comparing the results obtained for differentiation between compounds using simple correlation coefficient values from the 2D images and those achieved using Raman and LIBS spectra separately. Additionally, the effect of two spectral pretreatments (autoscaling and normalization) on the generation of the 2D image was evaluated. The results derived from this study demonstrate that the 2D image improves the identification of compounds, mainly in those critical situations in which it is not easy to differentiate them when Raman spectroscopy or LIBS is used separately.
Read full abstract