Three-dimensional Reconstruction and Measurement of Avian Eggs through Digital Imaging

This paper presents a computer vision-based method for the 3-D (three-dimensional) reconstruction and characterization of avian eggs. Two low-cost cameras are used to acquire images of eggs from top and side views, respectively. The image segmentation is performed using the image binarization technique. The contour-slice based method is employed for the 3-D reconstruction. The geometrical parameters of avian eggs, such as length, breadth, volume and surface area, are then computed based on the reconstructed model. The performance of the system is evaluated using eggs from different breeds and sizes. Comparative results between the physical measurement and the proposed approach suggest that the digital imaging approach has an overall accuracy of 98% for the geometrical parameter measurement of avian eggs.