Compact three-dimensional fluorescence spectroscopy and its application in food safety

Three-Dimensional fluorescence spectroscopy is an indispensable tool for identification of substances by virtue of its ability to supply abundant characteristic information and high spectral resolution. In this work, we present a compact three-dimensional fluorescence spectrometer utilizing 18 different wavelength LEDs as excitation light source, instead of the conventional bulky multi-wavelength illumination strategy. This spectrometer can in situ obtain the spectral information accurately and fast, while reducing the weight and volume of system. To show its broad biochemical utility, we utilize the compact spectrometer to measure four types of vegetable oil samples from different origins, and analyzed their spectral characteristics. Moreover, we construct a deep learning model based on the 2D-CNN network to detect sesame oil adulteration qualitatively and quantitively. The Compact fluorescence spectrometer proposed has the potential to classify and grade food oils in situ, addressing the problem of sesame oil adulteration on the market. •A compact three-dimensional fluorescence spectroscopy system was constructed.•The system can obtain three-dimensional fluorescence spectra of food in situ.•A deep learning model was developed to process in-situ spectral information.•Applications were executed for the quantitative analysis of sesame oil adulteration.