Pomegranate juices from different origins were served as the sample. Preprocessing method of near infrared spectroscopy data were determined and the feature were extracted by wavelet analysis. The three parameters of support vector machine (SVM) were optimized by genetic algorithms(GA).The model for determination of anthocyanin content in pomegranate juice by near infrared spectroscopy and support vector machine was established. The result showed that the model had the predication performance, the root mean square error of the validation and the determination coefficients were 0.019 766 and 0.999 2. Near-infrared spectroscopy technology could be used for the quantitative detection of anthocyanin content in pomegranate juice.

near infrared spectroscopy; pomegranate juice; anthocyanin; support vector machine