The method based on ATR-FTIR spectroscopy combined with feature variable selection for the boletus species and origins identification.

Abstract

Wild boletus mushrooms, which are macrofungi of the phylum Basidiomycetes, are a nutritious and unique natural food that is widely enjoyed. Since boletus are consumed with problems of indistinguishable toxic and non-toxic species and heavy metal enrichment, their species identification and traceability are crucial in ensuring quality and safety of consumption. In this study, the attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy technique combined with three feature variable extraction methods, manual selection method, semi-manual selection method, and algorithm method, were used to improve the accuracy and computational speed of the model identification, and the models were established for the identification of boletus species with an accuracy of up to 100% as well as for the identification of boletus origin with an accuracy of 86.36%. It was found that the best methods to improve the accuracy of the models were semi-manual selection, manual selection and algorithmic selection in that order. This study can provide rapid and accurate species identification and origin traceability of wild boletus, and provide theoretical basis for the rational use of feature variable selection methods.