Rapid, Minimally Invasive Prediction of Starch and Moisture Content in Saffron Corms Using Visible–Near-Infrared Spectroscopy Combined with Machine Learning

The starch and moisture content of saffron corms are critical indicators of their flowering potential and yield. This study investigated the use of rapid, minimally invasive VNIR reflectance spectroscopy measurement to assess these parameters. The measurements were used to develop predictive models through four machine learning algorithms (PLSR, RF, SVR, and GPR). Spectral data were obtained from 130 fresh corm samples. Wavelength analysis identified key starch-sensitive intervals (~930–1000 nm and ~1150–1220 nm) and a broad moisture-sensitive region (~900–1350 nm). Among the evaluated models, the combination of the multiplicative scatter correction pre-processing method and Gaussian process regression (MSC-GPR) demonstrated the optimal predictive performance for water content (R2 = 0.92, RMSE = 0.71%, RPD = 4.56, RPIQ = 5.37), and the combination of the MSC method and partial least squares regression (PLSR-MSC) demonstrated moderate performance for starch content (R2 = 0.73, RMSE = 28.7 mg g−1, RPD = 2.14, RPIQ = 2.81, dry weight). These results demonstrate the viability of VNIR spectroscopy as a minimally invasive tool for the pre-planting assessment of saffron corm quality under laboratory conditions. The method provides a laboratory-based framework for corm screening and selection, with potential for future adaptation to field settings using portable spectrometers following expanded calibrations and advanced modeling techniques.