UAV-Based Multispectral Maize Dataset for Water Stress and Rust Detection (Water Stress 2025 and Common Rust 2025) — Representative Subset (v1.0)

This dataset presents a representative subset of a UAV-based multispectral maize dataset developed for the detection of water stress and common rust disease under real field conditions. The dataset was acquired using unmanned aerial vehicles (UAVs) equipped with multispectral sensors, capturing high-resolution imagery across multiple spectral bands. The subset is curated to ensure transparency, reproducibility, and validation of the results reported in the associated scientific study. Each sample in the dataset includes:- RGB imagery in JPG format- Multispectral data in NPY format (6 spectral channels)- Pixel-wise ground truth masks in PNG format- Visualized annotation masks for qualitative inspection The dataset covers multiple semantic classes, including soil, healthy vegetation, low water stress, high water stress, and rust infection. The subset has been carefully selected to represent the diversity and distribution of the full dataset. This subset is publicly provided for research validation and reproducibility purposes. Due to the large volume of the complete dataset and its ongoing use in doctoral research, the full dataset will be made publicly available upon acceptance of the associated manuscript. This dataset is intended to support research in precision agriculture, plant disease detection, and multispectral image-based semantic segmentation.