NIRS-based detection advances in agriculture: Data enhancement, characteristic wavelength selection and modelling...

Near-infrared spectroscopy (NIRS) has gained increasing attention in agricultural element detection due to its advantages such as rapid analysis, low cost, and non-destructive measurement. With the continuous advancement of chemometric techniques, NIRS has become a powerful tool for extracting meaningful information from complex spectral data. Despite its growing applications, several critical challenges remain, particularly in enhancing data quality, selecting informative features, and developing robust modelling approaches. This review provides a comprehensive overview of recent progress in these areas, offering insights into data augmentation strategies, feature selection methods, and advanced modelling techniques. While the focus is on agricultural applications, the methodologies and conclusions presented herein are broadly applicable to other domains where NIRS plays a significant role.