Using dielectric properties and intelligent methods in separating of hatching eggs during incubation

Abstract

Fertility is an important trait in poultry industry so that any reduction in fertility resulted in a huge economic loss by rising incubation costs. In this study, dielectric constant and loss factor of eggs were used as a non-destructive, cheap, and precise method to identify dead embryos and infertile eggs, during incubation. For this purpose, artificial neural network (ANN) and support vector machine (SVM) classifiers were used. The result indicated that SVMs truly identified dead embryos with 100% accuracy at day 18 of incubation before hatching. In addition, SVM could also correctly identify un-hatching eggs, including dead embryos and infertile eggs with 92.31% accuracy at the 5th day. Neural network correctly classified un-hatching eggs with accuracy of 87.5% and 86.7%, respectively. Application of these two different classifiers showed that SVM yield better performance than ANN. The use of these capacitance properties not only present an automatic measurement method in detecting fertilized eggs in chicken, but also can apply in other strains of birds, which might be beneficial in birds’ hatchability system.