Multispectral reflectance dataset of medicinal Cannabis sativa cultivars for varietal classification

The dataset comprises 800 unique spectral signatures collected from four distinct cultivars: Embera CBD, Yocoto CBD, Limón CBD, and Valle de los Umbras. Data acquisition was performed using an Ocean Optics RED TIDE USB650 spectroradiometer with an optical fiber probe, covering a wavelength range from 410 nm to 890 nm. Measurements were taken under controlled ambient light conditions within a greenhouse, with cultivation parameters standardized. Spectral signatures were obtained from 30 plants of Embera CBD and Valle de los Umbras, and 20 plants of Limón CBD and Yocoto CBD varieties. Data collection spanned three distinct phenological stages (early vegetative, early flowering, and mid-flowering) and included measurements from both upper and lower leaves. Each signature represents an average of three measurements per leaf sample, calibrated with a white reflectance reference panel. Raw data were processed and normalized using Spectroscopy OceanView software, R, and JASP. The dataset is structured with columns for date (phenological stage), variety, plant ID, leaf location, and 2442 reflectance values corresponding to specific wavelengths. This publicly available dataset serves as a valuable resource for developing and training machine learning models for non-destructive and cost-effective classification of <i>Cannabis</i> varieties based on spectral characteristics, and for exploring correlations between spectral signatures and plant properties like cannabinoid content.