&Amp;lt;i>automatic Identification Method of Varieties of Pesticide Residues in Fruits</i>

Song Liu,Yankun Peng,Yongyu Li,Yan Lia,Kuanglin Chaob,Jianwei Qinb

DOI: https://doi.org/10.13031/aim.201900786

2019-01-01

Abstract:In recent years, the problem of pesticide residues has gradually become the biggest hidden danger to the safety of fruits and vegetables. The traditional pesticide detection method has the disadvantages of complicated pre-processing, long consumption time, expensive instrument, etc. It is not suitable for on-site and real-time detection. Raman spectroscopy technology has the unique advantages of simple, fast and sensitive, and can realize real-time detection.In this paper, the pesticides commonly used on the surface of apples, such as acetamiprid, chlorpyrifos, deltamethrin and prochloraz, were studied. The construction of several pesticide Raman spectroscopy libraries and their rapid identification of apple surface pesticides were discussed. The Raman spectras of four pesticides were collected five times using a Raman spectrometer, and the Raman spectra of each pesticide were averaged, and the spectral library was constructed after pre-processing such as shear smoothing and baseline calibration. Different concentrations and types of pesticide reagents were added to the surface of 96 identical samples, and Raman spectra of each sample to be tested were collected. Compared with the spectrum of the spectral library, feature extraction, recognition algorithms and similarity algorithms are used to qualitatively distinguish apple surface pesticide residue After comparison with the spectral library, there are 92 samples that can determine the pesticide residue on the surface of the apple. The accuracy rate is as high as 95.8%, which is highly feasible. Raman spectroscopy library is used for rapid qualitative identification of pesticide residues on fruits and vegetables, and has broad application prospects.