Method for rapidly identifying most dominant mold type of mildewed tobacco

The invention discloses a method for rapidly identifying the most dominant mold type of mildewed tobacco, which comprises the following steps: tabletting a mildewed tobacco leaf sample diluent by using a rapid mold yeast test piece, performing culturing, and characterizing bacterial colonies and feature metabolites thereof on the rapid mold yeast test piece by using near infrared spectroscopy to obtain a plurality of spectral data; preprocessing spectral data by using a discrete wavelet transform algorithm to obtain a plurality of wavelet coefficient reconstructed spectral data, and then identifying feature information of different types of most dominant fungi from the wavelet coefficient reconstructed spectral data by using a random forest algorithm so as to establish a most dominant strain identification model; and identifying the most dominant mold type of the mildewed tobacco through the most dominant strain identification model. The method has the advantages of being simple in pretreatment step, low in r