Method for lossless prediction of chemical components of eggplant sleeve tobacco leaves based on infrared spectrum technology

Fuente: WIPO "tobacco agriculture"
The invention belongs to the technical field of cigar tobacco content detection, and particularly relates to a method for non-destructively predicting chemical components of cigar sleeve tobacco leaves based on an infrared spectrum technology, which comprises the following steps: S1, collecting near infrared spectrum and the content of conventional chemical components; s2, preprocessing near infrared spectrum data; s3, grouping: grouping to obtain a training set, a test set and a check set; s4, establishing and evaluating a model; and S5, performing out-of-mold inspection on the model. The model established by the invention is used for detecting the conventional chemical components in the tobacco leaves, does not need to damage the tobacco leaves, is convenient to detect, can be used for large-scale determination of the content of the conventional chemical components in the tobacco leaves, can enter the market and is low in cost.