Abstract:A rapid detection method for reveratrol content in red wine was built by hyperspectral imaging technology combined with fluidized bed preconcentration technology. The reveratrol content in red wine was determined by high performance liquid chromatography. An enrichment apparatus of fluidized bed was built to explore the enrichment effects, which five macroreticular resins, such as HPD826, DA-201, AB-8, H103 and HPD600, enriched reveratrol in red wine. The result showed that H103 resin had the best enrichment effect. H103 resin was used to enrich resveratrol from wine, the 900~1700nm spectral images of resins which absorbed reveratrol were collected. Compared with five spectra preprocessing methods (MSC, SNV, SG-S, RN and QN), RN was selected as the best method for modeling effects of resveratrol content. Then, six regression models based on PLSR, SVMR (LK-SVMR, PK-SVMR, RBF-SVMR and S-SVMR) and PCR were established. Correction models based on LK-SVMR and PLSR were selected to evaluate their accuracy and stability in prediction sets. Finally, the PLSR model was chosen as the best model. The research showed that the quantitative prediction model of resveratrol in red wine based on PLSR obtained the best prediction effect, its RP was 0.8528, RMSEP was 0.0360, RC was 0.8783, and RMSEC was 0.0330. The results provided references for the detection of trace components by hyperspectral technology.