Non-Invasive Egg Monitoring Using Machine Vision Technology for Distinguishing Fertilized and Untilized Eggs

This study proposes a non-destructive method of measuring fertilized and non-fertilized eggs during the incubation process, using a machine vision system. The experiment involved 28 Rhode Island Red eggs, which were incubated in a basic incubator consisting of a heating and humidity chamber with automated temperature control. The study employed a machine vision system to measure the length and width of eggs and to monitor the incubation process by capturing daily images of the eggs from the beginning of incubation until hatching. Throughout the incubation period, egg weights and photographs were meticulously documented on a daily basis. The egg samples had an average length of 5.42 cm and an average width of 4.16 cm. Furthermore, on the first day of testing, the egg samples had an average weight of 54.06 g, an average volume of 49.24 µm³, and an average density of 1098.71 kg/m³. The study's findings showed a linear decrease in both egg weight and density as the days of incubation increased, for both fertilized and unfertilized eggs. Additionally, the machine vision system developed for this study can accurately track and measure fertilized eggs by analyzing the increased number of blood vessels within the egg. The fertilized eggs display a rapid increase in the number of blood vessels between days 3 and 7. Beyond day 7, the blood vessels area in the binary image of fertilized and hatchable eggs surpasses 200,000 pixels. Of the 26 eggs that hatched by day 21, only 7 successfully developed into chicks, resulting in a success rate of 27%. Starting from the fifth day of incubation, this approach can be utilized for non-destructive detection, allowing for differentiation between fertilized and unfertilized eggs.