抄録
We describe spectral classification of pesticide residue. This classification is done by using the attenuated total reflection method (ATR) of Fourier transform infrared spectroscopy (FT-IR). Recently, there has been increased concern about food safety. Not only consumers but also farmers are concerned with levels of pesticide residue. Furthermore, the dispersion of the pesticide (drift) has become an urgent problem. We used ATR Infrared spectroscopy to detect drift and to classify its. Spectral features have been extracted by using soft independent modeling of class analogy (SIMCA) appropriate to the concentration. The spectra of the calibration set were classified into each class on the basis of the concentration, and SIMCA was used to make the classification models. The concentration of pesticide residues of unknown samples was classified by applying the spectra to the classification model.