Pesticides are commonly used in farming to control agricultural pests and weeds. Excessive pesticide use can lead to severe problems such as contaminated produce, food poisoning, environmental pollution and soil contamination. A high volume of pesticide residue accumulation may also cause considerable damage to the environment and human health, and could contaminate surface water, groundwater and soil resulting in animal mortalities.

Thus, early detection and classification of pesticide residue could help consumers choose residue-free leafy vegetables. This research was performed to evaluate the performance of different classification methods to classify spectral data collected from 60 pesticide-free samples; for this research purpose, cabbage was selected. Deltamethrin pesticide was sprayed on the samples at different dilution concentrations namely pesticide-free (PF), pesticide-low (PL), pesticide-medium (PM) and pesticide-high (PH). The spectral data of the cabbages was recorded using a spectrometer with an effective wavelength in the range of 400 to 1000 nm. The concentration of the pesticide residues in each cabbage sample was quantified using gas chromatography with an electron detector (GC-ECD). Three classification methods investigated in this study were artificial neural networks (ANN), support vector machines (SVM) and logistic regression (LR). The results show that LR, SVM and ANN yielded excellent classification accuracy of 95, 88 and 87%, respectively. This study revealed that spectroscopic measurement coupled with classification methods are promising techniques for detecting and classifying pesticide residues in cabbage samples.

C.D.M. Iskandar, M.N. Nawi, R. Janius, N. Mazlan, T. Lin, L. Chen. Classification of Pesticide Residues in Cabbages Based on Spectral Data, AMA, Agricultural Mechanization in Asia, Africa and Latin America, 2021, Volume 52, Issue 03, 4033-4042

Date of Input: 24/11/2022 | Updated: 24/11/2022 | ainzubaidah