Data from: Hyperspectral image reconstruction for predicting chick embryo mortality towards advancing egg and hatchery industry

This dataset includes both ground-truth hyperspectral reflectance spectra and reconstructed spectral data generated from RGB images using various deep learning models. Fertile chicken eggs were imaged during incubation, and the corresponding embryo outcomes (e.g., viable or mortality) were also included. The ground-truth spectral data were acquired using a hyperspectral imaging (HSI) system in the visible and near-infrared (VNIR) range. Multiple deep learning models, including HRNET, EDSR, RESTORMER, and MST were trained to reconstruct hyperspectral spectra from RGB images. The reconstructed spectral data are included for model comparison and benchmarking. This dataset is intended for researchers working on hyperspectral image reconstruction, embryo viability prediction, and machine learning applications in precision poultry farming. It enables further development and validation of models for non-invasive embryo assessment and supports innovation in automated hatchery management.