HSI-PVL-KZ

This dataset contains laboratory-acquired hyperspectral images of spring wheat (Triticum aestivum L.) collected across multiple phenological stages during the 2025 growing season. The data were acquired using a FigSpec FS-13 hyperspectral camera in the spectral range of 400–1000 nm under controlled illumination conditions.The dataset includes hyperspectral cubes with corresponding metadata describing acquisition dates, locations (district and field), phenological stages (Zadoks scale), and plant health status. The samples represent healthy plants as well as multiple disease categories, including septoria leaf blotch, root rot, rust, yellow leaf symptoms, brown rust, loose smut, and mixed infections.For selected samples, regions of interest (ROIs) were annotated at the organ level (leaf, ear, background) and used to extract mean spectral signatures and vegetation indices such as NDVI and red-edge-based features. The dataset also includes dry plant samples from a 2024 reference collection, enabling comparative analysis of spectral changes associated with moisture loss.The primary purpose of this dataset is to support research in hyperspectral data analysis, plant disease monitoring, and machine learning–based classification. It is suitable for benchmarking traditional classifiers (e.g., Random Forest, Maximum Entropy, SIMCA) as well as for developing and validating deep learning approaches.A subset of the dataset is publicly available for exploratory use. Access to the full dataset and detailed annotations can be obtained by contacting the authors.