Qualitative Identification of Pesticide Residues in Pakchoi Based on Near Infrared Spectroscopy

Abstract

This paper provided a new method for rapid and nondestructive classification and identification of vegetable residues. Near Infrared Spectroscopy (NIR) of two kinds of pakchoi were collected. One kind of pakchoi was non-pesticide. And the other was low concentration of pesticide residues . In this paper, wavelet soft threshold was used to pre-treat the spectral data. Then Principal Component Analysis (PCA) was run to reduce the dimensionality of spectral data. Next, Linear Discriminant Analysis (LDA) and K-Nearest Neighbor Classification and Recognition (KNN) was run to classify and discriminate. The experimental results show that the prediction rate is 95%. This method is accurate and effective.