"Hyperspectral Imaging (HSI) of Corroded Steel and Iron with Different Coatings"

"This dataset presents hyperspectral images designed for the study and characterization of corrosion on metallic surfaces coated with different types of paint. The dataset consists of two subsets acquired under controlled laboratory conditions and radiometrically calibrated using white and black reference targets, ensuring normalized reflectance values between 0 and 1. The images were captured using the Specim FX10 (400\u20131000 nm) and Specim FX17 (900\u20131700 nm) hyperspectral cameras, corresponding to the Visible and Near-Infrared (VNIR) and Short-Wave Infrared (SWIR) spectral ranges, respectively. The first subset includes 45 iron samples with 99.60% purity, distributed across five coating types: nitrosynthetic yellow, nitrosynthetic brown, acrylic yellow, acrylic brown, and uncoated metal. For each coating type, nine plates were prepared, divided into three corrosion levels (increasing degrees of corrosion) and an additional Level 0, corresponding to the initial condition of the level 1 samples before exposure to the salt spray chamber. Corrosion was induced through accelerated testing at 35 \u00b0C with a 5% NaCl solution, and each sample was imaged in both VNIR and SWIR ranges. The second subset comprises nine varied unknown samples collected from industrial environments, composed of iron or steel of uncertain alloy composition, and captured following the same calibration procedure. In total, the dataset comprises 45 laboratory plates and nine industrial samples, providing a comprehensive hyperspectral dataset for validating spectral analysis methodologies and developing machine learning models for corrosion detection and surface classification of metallic materials."