Raw Multispectral Image Dataset for Sugarcane Disease Severity Assessment

This dataset contains raw multispectral images of sugarcane leaves collected under field conditions in Yunnan Province, China. The dataset was established for sugarcane disease severity assessment and includes five spectral bands: RGB, green, red, red-edge, and near-infrared (NIR). The images were collected from two experimental sites: (1) Yunnan Provincial Sugarcane Germplasm Resource Nursery; (2) Gengma Sugarcane Plantation Base, Yunnan Province, China. The multispectral images were acquired using a multispectral camera equipped with green (560 nm ±16 nm), red (650 nm ±16 nm), red-edge (730 nm ±16 nm), and near-infrared (860 nm ±26 nm) bands. The dataset contains sugarcane leaves with different disease severity levels, including mild, moderate, moderately severe, and severe symptoms. Each sample was recorded using multiple spectral bands. Images with the same sample ID correspond to the same leaf sample captured in different spectral bands. This dataset can support research in multispectral image analysis, plant disease recognition, disease severity classification, spectral feature extraction, and precision agriculture.