Research On Nondestructive Detection Technology Of Pesticide Residues On Cauliflower Surface Based On Hyperspectral Technology

Cauliflower is one of the main vegetables in the "vegetable basket project".In the cultivation and production of cauliflower,it is necessary to rely on the application of pesticides to increase yield and reduce the occurrence of diseases and insect pests.If pesticides are used at the wrong time or excessive use of pesticides,it will lead to the problem of excessive pesticide residues in dishes.This problem not only leads to the destruction of ecology and environment,but also threatens people's food safety and life and health.Facing people's life and health is an important part of General Secretary Xi Jinping's "four orientations" for the development of science and technology.As a key link to protect people's life and health,food safety and nutrition and health are important livelihood projects and popular projects.Therefore,exploring efficient and rapid pesticide residue detection methods has become one of the urgent problems to be solved in this field.In this paper,four pesticides on cauliflower are taken as the research objects,and the qualitative analysis of these four pesticides is carried out based on hyperspectral technology,and the feasibility of rapid and non-destructive detection of pesticide residues by hyperspectral technology is studied.Based on the feasibility of the pesticide dispersal law,the pesticide was tracked and detected by chromatographic method.The main research contents and results are as follows:(1)Hyperspectral imaging technology was used to conduct non-destructive testing of bacillus thuringiensis,beta-cypermethrin and indoxacarb respectively.One of the best results of the detection was tracked and studied to observe the safe interval for pesticides.After hyperspectral imaging was performed on cauliflower samples containing and without pesticides,the spectral data of the region of interest was extracted.The 20 bands before and after the original spectral data were eliminated to reduce the influence of noise.Convolution smoothing method(S-G),multivariate scattering correction(MSC)and standard normal variate(SNV)algorithms were used to optimize the remaining 216 bands(950 nm?1666 nm),respectively.The characteristic wavelength of the three pesticide spectral data was extracted using the competitive adaptive reweighted(CARS)algorithm to improve the discriminant operation speed,and finally the partial least squares(PLS)discriminant model was established.The PLS model optimized based on SNV had the highest recognition accuracy of the three pesticides on cauliflowers.Among them,the test results of the pesticide samples of indoxacarb were the best,with a recognition rate of 100%,and the 7-day detection result of this pesticide was consistent dissipation law of pesticides.(2)Spectral data for regions of interest were extracted after hyperspectral imaging of cauliflower samples with and without carbaryl indocarb.After processing and analyzing the spectral data of the pesticide according to the above method,the pesticide is tracked and detected by combining the hyperspectral technology and the chromatographic method to verify the accuracy of the hyperspectral technology.The results showed that the SNV-optimized PLS model had the highest accuracy in determining whether the cauliflower contains carbaryl indocarb pesticides,and the residue rate of indoxacarb,the main component of this pesticides,was determined based on hyperspectral technology.The detection is similar to the actual residue rate of this pesticide,and the error between the estimated half-life and the actual detection result is only 0.14 d.