Automated system for extraction of eggshell morphometric characteristics in chicken eggs

The evaluation of eggshell morphometric characteristics is essential to ensure the quality of eggs intended for both consumption and incubation. Conventional methods for this analysis are mostly manual, destructive, time-consuming, and prone to human error. In this context, the present study proposes the development of an automated system based on computer vision for extracting key eggshell features: maximum length and width, area, porosity, and cracks. The system was designed to capture images from two positions, enabling high-resolution analysis of overall egg morphometry and shell porosity. A total of 326 commercial eggs from different production systems were analyzed, including 162 brown-shelled and 164 white-shelled eggs. The results showed a high correlation between the values obtained by the automated system and those from manual measurement methods, with coefficients of determination (R²) above 0.85 for the morphometric variables. Porosity analysis showed better performance for white-shelled eggs (R² = 0.97) compared to brown-shelled eggs (R² = 0.87). Crack detection achieved an accuracy above 90 % for both shell types, demonstrating the effectiveness of the automated method. It is concluded that the proposed system is a promising tool for application in the poultry industry, providing greater efficiency, precision, and reliability in eggshell quality assessment.