Author(s)

Yan Cai, Dianxu Ma, Xiaopan Li

Corresponding Author

Yan Cai

Abstract

This study explores a method for rapid detection of the origin of Green Sichuan Pepper based on near-infrared spectroscopy. A total of 260 samples of peppercorns from 5 producing areas including Longtoushan Town in Ludian County, Suoshan Town in Ludian County, Xiaozhai Town in Ludian County, Jiangdi Town in Ludian County and Tianba Town in Zhaoyang District were collected. Spectra are collected. The original spectra were preprocessed by methods such as wavelet decomposition and denoising, and then radical basic function (RBF), support vector machine (SVM), and partial least squares (PLS) were used to establish the origin identification model. The research shows that the identification accuracy of SVM and RBF neural network models are significantly improved, up to 100%; wavelet decomposition denoising and baseline correction can significantly improve the accuracy of the Green Sichuan Pepper identification model. The rapid detection method of Green Sichuan Pepper based on near-infrared spectroscopy is feasible.

Keywords

Near-infrared spectroscopy, Green Sichuan pepper, Origin identification, Pretreatment, Rapid detection