Research on Identification of Boletus tomentipes with Different Storage Period by UV and FT-IR Combined with Data Fusion
Article
Figures
Metrics
Preview PDF
Reference
Related
Cited by
Materials
Abstract:
Long term storage of wild edible mushrooms would cause microbial proliferation and physico-chemical changes, affecting the quality. In order to ensure the security and quality, it was essential to establish a quick and efficient method to identify wild edible mushrooms with different storage periods. Ultraviolet (UV) and fourier transform infrared (FT-IR) spectra of 77 fruit bodies of B. tomentipes (5 years of storage) were preprocessed by using Savitzky-Golay (SG) smoothing, second derivative (2-D) and standard normal variate (SNV), and UV, FT-IR, low-level and mid-level data fusion models were established with partial least squares discriminant analysis (PLS-DA). The results showed that the optimal pretreatment of UV and FT-IR spectra were SG+2-D and SG+2-D+SNV, respectively, and the classified individual errors were 10, 6, 4 and 3 in UV, FT-IR, low-level and mid-level data fusion models.Theaverage of R2cal in mid-level data fusion model was closest to 1, and the average of RMSECV was minimum , which indicated that the effects of mid-level data fusion model were better than those of other three models. The UV and FT-IR mid-level data fusion strategy combined with PLS-DA could accurately identify the B. tomentipes with different storage periods, which provided a novel reference for quality evaluation of wild edible mushrooms.
