DETECTING FERTILITY OF HATCHING EGGS USING MACHINE VISION. I: HISTOGRAM CHARACTERIZATION METHOD

The ability to automatically detect fertile eggs during early incubation would allow timely removal of dead embryos and infertile eggs from incubation thereby contributing to overall hatchery profits. A method using machine vision for detecting well developing eggs during the third and fourth day of incubation was developed. Images of eggs were acquired using backlighting with a high intensity candling lamp. The images were enhanced using a histogram stretch and parameters were computed to describe the shape of the histogram of the resulting image. An algorithm was developed to distinguish fertile from infertile eggs based on histogram shape parameters. The algorithm performed with accuracies of 96 to 100% at day four of incubation and 88 to 90% at day three of incubation.