Deep Potato - hyperspectral imagery of potato leaves with reference measurements dataset: towards potato physiological features modeling. The dataset contains 8 high-resolution hyperspectral orthophoto maps for deep learning modeling, stored in the "orthophoto_maps" folder. The maps are encoded in ENVI format. For each map 2 files: HDR file with the necessary information and BSQ file with imagery data are provided. The BSQ files are compressed as ZIP. - for reference, there are 60 lab measurements supplemented as well as 74 physiological measurements based on collected plants' material, both with precise geolocation data linking the measurements and the orthophoto maps. Those files are stored in the "measurements" folder, - hyperspectral images, extracted from orthophoto maps, stored in the "images" folder. - a "dataset_presentation.html" file with a brief presentation of the dataset.