基于可见/近红外高光谱成像技术的番茄表面农药残留无损检测

Detection of pesticide residue on cherry tomatoes by hyperspectral image technology

以番茄为研究对象，应用可见／近红外高光谱成像技术对水果表面农药残留的无损检测研究。用蒸馏水将嘧霉胺农药稀释成1：20，1：100，1：5003个梯度，将不同浓度的溶液分别滴到60个洗净的番茄表面，形成3×3矩阵。放置在通风阴凉处12h后，应用高光谱系统（400～1000nm）采集光谱图像信息。利用主成分分析法获得主成分图像（PC），并根据第二主成分图像（PD2）的权重系数选取特征波长564，809，967nm。采用波段比（564nm／809nm）结合适当的图像处理方法对番茄表面的农药残留进行检测。高浓度（1：20，1：100）农药点检测率为100％，而低浓度（1：500）农药点的检测率为0。结果表明，高光谱成像技术对高浓度农药残留具有较好的检测效果。

Based the cherry tomatoes as the research object, applied the visible near infeared hyperspectral image technology nondestructive to test the pesticide residues on fruit surface. The pesticide was diluted to 1 ： 20,1 ： 100 and 1 ： 500 solution with distilled water. Then a 3× 3 matrix of dilutions was applied to 60 cleaned sample with three different solutions. After 12 h, the heperspectral system was used to obtain the images between 400 nm and 1 000 nm. The principal component （PC,） images ware achieved by principal compo nent analysis （PCA）. The characteristic wavelengths of 564 nm, 809 nm, 967 nm were achieved by the second principal component images＇ weight coefficient, wave band （564 nm/809 nm） with proper image processing detectrf the pesticide residues on tomatoes surface. The high concentration of pesticide（1 ： 20, 1 ： 100） detection rate was 100%, while, the low concentration（1 ： 500） was 0. The research showed that the heperspeetral image technology could be used to effectively detect pesticide residue on fruit surface.