Abstract:Aiming at the problem of pesticide residues in vegetables sold in the market, an efficient and nondestructive method for the qualitative classification and identification of pesticide residues in Chinese cabbage is proposed. Three groups of Chinese cabbage leaves and cyhalothrin are used as the research objects. Two groups of Chinese cabbage are sprayed with two different concentrations of pesticides (the ratios of pesticide to water are 1∶500 and 1∶20 respectively), and three types of samples are formed, which contain no pesticides, mild pesticide residues and severe pesticide residues. Three kinds of samples are collected by near infrared spectroscopy, and the spectral data are preprocessed by wavelet soft threshold, then the dimension is reduced by principal component analysis. Fisher decision and K-nearest neighbor classification are performed as well. The experimental results show that the correct identification rate of the two kinds of samples without pesticide residues and with mild pesticide residues is 95%, and that of the samples with mild and severe pesticide residues is 90%, which proves that this method is effective for the qualitative classification and identification of pesticide residues in Chinese cabbage and provides a new way of thinking for the qualitative classification and identification of pesticide residues.