Hyperspectral imaging dataset for non-destructive fertility and structural evaluation of chicken eggs

Hyperspectral imaging (HSI) is a smart, non-destructive sensing that integrates spectral and spatial information, advancing fertility prediction and structural evaluation of eggs. Despite increasing interest in poultry research, there remains no publicly accessible HSI dataset, limiting the development and validation of robust, generalizable machine learning models. To address this gap, the present study provides a comprehensive HSI dataset encompassing key egg parameters, pre-incubation fertility status, eggshell thickness, yolk mass, eggshell strength, and associated morphological parameters (intact egg mass, major and minor diameters). The dataset consists of hyperspectral images of 1228 white Leghorn chicken eggs acquired using a line-scan transmission HSI system (374-1015 nm). Each egg is accompanied by validated reference measurements, enabling supervised learning tasks such as regression and classification. Raw hyperspectral cubes (.mat format) and annotated spectral metadata (.csv format) are structured for easy access and reuse. Rigorous technical validation confirmed the dataset's reliability. This open-access resource is designed to accelerate precision poultry research and promote the development of non-invasive, data-driven egg evaluation and quality assurance techniques.