Dataset and code for multimodal spectral-imaging monitoring of Gardeniae fructus extract drying

This dataset contains the data and analysis code associated with the study “Multimodal spectral-imaging fusion with machine learning for monitoring Gardeniae fructus extract drying”. The deposit includes cropped RGB region-of-interest images used for image-feature extraction, hyperspectral reflectance source workbooks, moisture reference measurements, HPLC-based reference measurements of major bioactive components, processed RGB feature tables, image-to-sample mapping tables, and Python scripts for RGB feature extraction, spectral-index screening, and baseline regression modelling. The full-size original photographs are not included; the public image data consist of the cropped RGB ROIs used in the analysis. Moisture reference values are available for 216 samples. HPLC component reference values are available for 215 samples; the unavailable component record is left blank in the processed tables and should be excluded from component-specific modelling.