Light absorption gas sensing technology has the characteristics of massive parallelism, cross-sensitivity and extensive responsiveness, which make it suitable for the sensing task of an electronic nose (e-nose). With the performance of hyperspectral resolution, spatial heterodyne spectrometer (SHS) can present absorption spectra of the gas in the form of a two dimensional (2D) interferogram which facilitates the analysis of gases with mature image processing techniques. Therefore, a visual e-nose system based on SHS was proposed. Firstly, a theoretical model of the visual e-nose system was constructed and its visual maps were obtained by an experiment. Then the local binary pattern (LBP) and Gray-Level Co-occurrence Matrix (GLCM) were used for feature extraction. Finally, classification algorithms based on distance similarity (Correlation coefficient (CC); Euclidean distance to centroids (EDC)) were chosen to carry on pattern recognition analysis to verify the feasibility of the visual e-nose system.
Keywords: SHS; feature extraction; gas sensing; pattern recognition; visual e-nose.